Upgrading of High-Aluminum Hematite-Limonite Ore by High Temperature Reduction-Wet Magnetic Separation Process

Transcription

1 metals Article Upgrading of High-Aluminum Hematite-Limonite Ore by High Temperature Reduction-Wet Magnetic Separation Process Xianlin Zhou *, Deqing Zhu, Jian Pan, Yanhong Luo and Xinqi Liu School of Mineral Processing and Bioengineering, Central South University, Changsha 483, China; (D.Z.); (J.P.); (Y.L.); (X.L.) * Correspondence: xlzhou_csu@csu.edu.cn; Tel./Fax: Academic Editor: Hugo F. Lopez Received: 8 January 2016; Accepted: 3 March 2016; Published: 8 March 2016 Abstract: The huge consumption of ores in China has attracted much attention to utilizing low grade complex resources, such as high-aluminum hematite-limonite ore, which is a refractory resource and difficult to upgrade by traditional physical concentration processes due to superfine size and close dissemination of minerals gangue minerals. An innovative technology for a high temperature reduction-magnetic separation process was studied to upgrade a high-aluminum ore assaying 41.92% Fe total, 13.74% Al 2 O 3 and 13.96% SiO 2. The optimized results show that final metal powder, assaying.46% Fe total, was manufactured at an overall recovery of.25% under conditions as follows: balling high aluminum ore 15% coal blended and at 0.3 basicity, reducing dried pellets at 1350 C for 25 min a total C/Fe ratio of 1.0, grinding reduced pellets up to 95%, passing at mm and magnetically separating ground product in a Davis Tube at a 0.10-T magnetic field intensity. The metal powder can be used as burden for an electric arc furnace (EAF). Meanwhile, nonmagnetic tailing is suitable to produce ceramic, which mainly consists of anorthite and corundum. An efficient way has been found to utilize high-aluminum resources. Keywords: high-aluminum hematite-limonite ore; aluminum- separation; high temperature reduction; wet magnetic separation 1. Introduction The growth of Chinese steel industry has been pushed forward by its rapid economic development in recent years, and 1126 million tons of ore have been consumed in 2014, where 933 million tons were imported according to data from Worldsteel Association [1]. However, resources are limited, and reserves of high-grade ores are declining, so it is of great importance to optimize mining rate of resources [2] and to utilize low-grade refractory resources, such as high-aluminum limonite ores. High-aluminum limonite ore is a typical refractory resource; it is difficult to upgrade by traditional processes due to superfine size and close dissemination of minerals gangue minerals [3,4]. Some high-aluminum ores were directly blended in a sintering mixture at a low ratio, but sintering process and properties of product sinter were adversely affected, which was reviewed by Lu et al. [5]. Meanwhile, high-aluminum ore affects viscosity and desulfurizing capacity of blast furnace slag [6]. Many upgrading approaches have been published for high-aluminum limonite ore, including physical and chemical processes [4]. Recent research was focused on gravity concentration and magnetic separation [7], flotation [8,9], sodium roasting and magnetic separation process [3,10 15]. However, and aluminum cannot be separated completely by physical processes, Metals 2016, 6, 57; doi:10.33/met

2 Metals 2016, 6, 57 2 of 12 Metals 2016, 6, x 2 of 12 so recovery is low by flotation process (around 70%), and as high as beyond 10% of sodium additive completely is needed by physical by processes, sodium roasting so process, recovery which is low brings by flotation risk of process degradation (around of 70%), refractory and materials as high as inbeyond furnaces 10% [16,17] of sodium and higher additive costs. is needed by sodium roasting process, which brings risk Theof coal-based degradation direct reduction-magnetic of refractory materials separation in furnaces process [16,17] is and efficient higher costs. way to recover fromthe low-grade coal based complex direct reduction magnetic ores [18] and secondary separation resources process containing is an efficient ferrous way material, to recover such as copper from low grade slag [19]. complex However, aores low reduction rate is a problem of reduction process, which is always operated under C [18] and secondary resources containing ferrous material, such as copper slag [19]. However, a low reduction for more than rate is 60a min. problem Elevating of reduction process, temperature which is an is effective always operated means tounder promote reduction C for rate more ofthan minerals. 60 min. Elevating Therefore, a reduction novel technology temperature of high is temperature an effective means reduction-wet to promote magnetic reduction separation rate of of high-aluminum minerals. limonite Therefore, is studied a novel in technology this paper. of high temperature reduction wet magnetic separation of high aluminum limonite is studied in this 2. Materials and Methods paper Raw Materials 2. Materials and Methods The chemical compositions of high-aluminum hematite-limonite ore used in this study are tabulated 2.1. Raw Materials in Table 1. The total grade of ore sample is 41.92%, and high alumina and silica contents were observed, assaying 13.74% Al 2 O 3 and 13.96% SiO 2, respectively. Combining The chemical compositions of high aluminum hematite limonite ore used in this study are X-ray diffraction (XRD) results shown in 1, minerals mainly consist of hematite and tabulated in Table 1. The total grade of ore sample is 41.92%, and high alumina and silica goethite, and aluminum mineral is kaolinite. s 2 and 3 illustrate representative SEM-EDS contents were observed, assaying 13.74% Al2O3 and 13.96% SiO2, respectively. Combining and optical micrographs of ore samples, which indicate that hematite is closely included kaolinite, X ray diffraction (XRD) results shown in 1, minerals mainly consist of hematite and leading to significant difficulties in separation of and aluminum. goethite, and aluminum mineral is kaolinite. s 2 and 3 illustrate representative SEM EDS and optical micrographs Table 1. Chemical of ore compositions samples, of which high-aluminum indicate hematite-limonite that hematite is ore closely (wt. %). included kaolinite, leading to significant difficulties in separation of and aluminum. Chemical Compositions (wt. %) Table 1. Chemical compositions of high aluminum hematite limonite ore (wt%). Fe total FeO Al 2 O 3 SiO 2 CaO MgO Mn K 2 O Na 2 O P S Pb Zn LOI Fetotal FeO Al2O3 SiO2 CaO MgO Mn K2O Na2O P S Pb Zn LOI Note: LOI, loss on ignition Note: LOI, loss on ignition. Intensity (Counts) - Hematite - Kaolinite - Goethite CuKα Two-Theta (deg) 1. XRD pattern of high aluminum hematite limonite ore. 1. XRD pattern of high-aluminum hematite-limonite ore.

3 Metals 2016, 6, 57 3 of 12 Metals 2016, 6, 57 3 of 12 Metals 2016, 6, 57 3 of Representative SEM BSE micrographs of high aluminum hematite limonite ore. (BSE, Back 2. Representative SEM-BSE micrographs of high-aluminum hematite-limonite ore. (BSE, Back Scattered 2. Electron Representative Imaging; SEM BSE H, hematite; micrographs K, kaolinite). of high aluminum hematite limonite ore. (BSE, Back Scattered Electron Imaging; H, hematite; K, kaolinite). Scattered Electron Imaging; H, hematite; K, kaolinite) Granular Granular hematite hematite (H) (H) filled filled in in particles of of kaolinite kaolinite (K) (K) under under reflected reflected light. light. 3. Granular hematite (H) filled in particles of kaolinite (K) under reflected light. CaCO3 CaCO3 of of analytical grade was used to adjust binary basicity of of blends. The soft coal was as a of on an air dry basis CaCO The soft coal was used a reductant agent, fixed carbon of 52.12% on an air dry basis 3 of analytical grade was used to adjust binary basicity of blends. (FCad), volatile matter of of 30.41% on a dry ash free (Vdaf) basis, ash of of 4.49% on on an an air air dry dry basis basis The soft coal was used as a reductant agent, fixed carbon of 52.12% on an air dry basis (Aad), (Aad), 0.58% of of S and S and a a melting temperature of 1376 C. Coal ash possesses high silica, calcium oxide oxide (FCad), and and volatile sulfur sulfur matter content, of as as 30.41% listed in in on Table a dry Soft ash coal free as (Vdaf) a reductant basis, is ash is % of passing 4.49% on at at 5 anmm, 5 air mm, and dry and some basis some (Aad), 0.58% soft of soft coal S and coal used used a melting for for blending temperature in in pellet of feed 1376 was C. milled Coalto ash % possesses passing at high at silica, mm. mm. calcium oxide and sulfur content, as listed in Table 2. Soft coal as a reductant is % passing at 5 mm, and some soft coal Table Table Chemical compositions of of coal coal ash ash (wt. (wt. %). used for blending in pellet feed was milled to % passing at %). mm. Chemical Chemical Compositions Compositions (wt. (wt. %) %) Fetotal Fe2O3 Table SiO2 2. Chemical Al2O3 compositions CaO MgO of coal ash K2O (wt. Na2O %). P S Fetotal Fe2O3 SiO2 Al2O3 CaO MgO K2O Na2O P S Chemical Compositions (wt. %) Fe total Fe 2 O 3 SiO 2 Al 2 O 3 CaO MgO K 2 O Na 2 O P S

4 Metals 2016, 6, 57 4 of Experimental Procedures The experimental flow sheet includes procedures as follows: mixing ore CaCO 3 and soft coal, pelletizing of mixture, high temperature reduction roasting of dried pellets, grinding reduced pellets and low intensity magnetic separation of ground product to manufacture metal powder. Mixtures were prepared by mixing ore, CaCO 3 and soft coal fines. The of CaCO 3 was determined by binary basicity, and soft coal blended in pellets was calculated by coal-bearing ratio multiplied by of ore sample. Then, green balls were made by balling mixtures in a disc pelletizer of 0.8 m in diameter and a 0.2 m rim depth, rotating at 38 rpm and being inclined at 47 to horizontal. The screened green balls of 8 16 mm were loaded into drying oven to dry at 105 C for 2 h until weight was unchanged. The dried pellets were put into corundum crucible and covered by some soft coal, where of soft coal was determined by C/Fe ratio. The crucible was covered by a graphite cover and put into muffle furnace (model: KSY-12-18, The Great Wall Furnace, Changsha, China), while reducing temperature was elevated to target value. When reduction time was ended, hot crucible was taken out and covered by pulverized coal to cool down quickly in air to prevent reduced pellets from being re-oxidized. The magnetic separation was performed in Davis Tube (model: XCGS-73, Changsha Research Institute of Mining and Metallurgy Co., Changsha, China) after 20 g of reduced pellets were ground in cone ball mill (model: RK/ZQMØ160ˆ60, Wuhan Rock Crush & Grand Equipment Manufacture Co., Wuhan, China). The metal powder was obtained after magnetic separation. The grinding fineness of reduced pellets was about 95% passing at mm, and magnetic field intensity was 0.10 T. The chemical compositions of ore and metal powder were measured by X-ray fluorescence spectroscopy (XRF, PANalytical Axios max, PANalytical B.V., Almelo, The Nerlands). The crystalline phase composition of material was determined using an X-ray diffractometer (XRD, D/Max-2500, Rigaku Co., Tokyo, Japan). Proximate analysis of coal and determination of fusibility of coal ash were conducted by Chinese standards GB/T and GB/T , respectively. Microstructures of reduced pellets were observed by scanning electron microscope (SEM, FEI Quanta-200, FEI Company, GG Eindhoven, The Nerlands), and compositional analyses were carried out using an energy dispersion system (EDAX-TSL, Ametek Inc., Paoli, CO, USA) in SEM Thermodynamic Analysis of Reduction Process Equations (1) (18) show possible reactions during reduction process of high-aluminum hematite-limonite ore, and FactSage7.0 (Thermfact/CRCT, Montreal, QC, Canada; GTT-Technologies, Herzogenrath, Germany) is applied to calculate Gibbs energy changes of all of reactions; results are depicted in s 4 and 5. In 4, line of Equation (1) is intersected by lines of Equation (3) and Equation (4) at T 1 and T 2. Because of Boudouard reaction (Equation (1)), CO concentration is higher than that of Equations (2) (4) at a temperature over T 2, so metallic begins to appear (Equation (4)) where CO concentration is over 58%. During T 1 to T 2, Fe 3 O 4 is reduced to FeO (Equation (3)), and CO concentration is between 42% and 58%. The region of T 1 < T < T 2 is stable zone of FeO phase. When temperature is lower than T 1, CO concentration is still higher than that of Equation (2), so Fe 2 O 3 is reduced to Fe 3 O 4 (Equation (2)). Consequently, regions of T < T 1, T 1 < T < T 2 and T > T 2 are equilibrium stable fields of Fe 3 O 4, FeO and Fe, respectively. Theoretically, reduction of oxide by solid carbon follows steps as Fe 2 O 3 ÑFe 3 O 4 ÑFeOÑFe beyond T 2 ; transformations Fe 2 O 3 ÑFe 3 O 4 and FeÑFeOÑFe 3 O 4 will occur when reduction temperature is lower than T 1 ; while Fe 2 O 3 ÑFe 3 O 4 ÑFeO and FeÑFeO are steps between T 1 and T 2 [20]. It is demonstrated that oxide is reduced by stepwise deoxidation, and wüstite (FeO) is an intermediate product. Because of presences of SiO 2 and Al 2 O 3,

5 Metals 2016, 6, 57 5 of 12 wüstite can easily chemically combine m to form fayalite (Fe 2 SiO 4 ; refer to Equation (6)) and hercynite (FeAl 2 O 4 ; refer to Equation (7)). Meanwhile, mullite (Al 6 Si 2 O 13 ) may occur by reaction between SiO 2 and Al 2 O 3 (Equation (8)) and n is degraded by FeO to form fayalite and hercynite (Equations (9) and (10)), both of which are unable to be reduced by CO rmodynamically [19,21]. Thus, Metals solid 2016, carbon 6, x is used as reductant, where fayalite and hercynite can be reduced5 of beyond 12 5 and 830 C, respectively. Moreover, alkaline-earth oxide can improve reduction of fayalite and between SiO2 and Al2O3 (Equation (8)) and n is degraded by FeO to form fayalite and hercynite hercynite by reacting SiO 2 and Al 2 O 3 chemically to form larnite (Ca 2 SiO 4 ) and calcium aluminate (Equations (9) and (10)), both of which are unable to be reduced by CO rmodynamically [19,21]. (CaAl Thus, 2 O 4 ), solid respectively. carbon is used as reductant, where fayalite and hercynite can be reduced beyond 5 and 830 C, respectively. Moreover, alkaline earth C ` CO 2 oxide 2CO can improve reduction of fayalite and (1) hercynite by reacting SiO2 and Al2O3 chemically to form larnite 3Fe 2 O 3 ` CO 2Fe 3 O 4 ` CO 2 T ą 570 C (Ca2SiO4) and calcium aluminate (CaAl2O4), respectively. (2) Fe 3 O 4 ` CO C + 3FeO ` CO 2 T ą 570 C CO2 = 2CO (1) (3) FeO ` CO Fe ` CO 2 T ą C 3Fe2O3 + CO = 2Fe3O4 + CO2 T >570 C (2) (4) 1{4Fe 3 O 4 ` 3{4Fe ` CO 2 T ă 570 C Fe3O4 + CO = 3FeO + CO2 T >570 C (3) (5) FeO + CO = Fe + CO2 T >570 C 2FeO ` SiO 2 Fe 2 SiO 4 (4) (6) 1/4Fe3O4 + CO = 3/4Fe + CO2 T <570 C FeO ` Al 2 O 3 FeAl 2 O 4 (5) (7) 2FeO + SiO2 = Fe2SiO4 (6) 3Al 2 O 3 ` 2SiO 2 Al 6 Si 2 O 13 FeO + Al2O3 = FeAl2O4 (7) (8) 4FeO ` Al 6 Si 2 O 13 2Fe 2 SiO 4 ` 3Al 2 O 3Al2O3+2SiO2= Al6Si2O13 3 (8) (9) 3FeO4FeO ` Al + 6 Al6Si2O13 Si 2 O = 2Fe2SiO4 3FeAl 2 O+ 4 3Al2O3 ` 2SiO 2 (9) (10) Fe 2 SiO 3FeO 4 `+ Al6Si2O13 2CO = 2Fe 3FeAl2O4 ` SiO+ 2 2SiO2 ` 2CO 2 (10) (11) Fe2SiO4 FeAl 2 O 4 ` + CO 2CO = 2Fe `+ SiO2 Al 2 O+ 2CO2 3 ` CO 2 (11) (12) FeAl2O4 + CO = Fe + Al2O3 + CO2 Fe 2 SiO 4 ` 2C 2Fe ` SiO 2 ` 2CO (12) (13) Fe2SiO4 + 2C = 2Fe + SiO2 + 2CO (13) FeAl 2 O 4 ` C Fe ` Al 2 O 3 ` CO FeAl2O4 + C = Fe + Al2O3 + CO (14) (14) Fe 2 SiO 4 ` 2CaO ` 2CO 2Fe ` Ca 2 SiO 4 ` 2CO 2 Fe2SiO4 + 2CaO + 2CO = 2Fe +Ca2SiO4 + 2CO2 (15) (15) FeAl 2 O 4 ` CaO ` CO Fe ` CaAl 2 O 4 ` CO 2 (16) FeAl2O4 + CaO + CO = Fe + CaAl2O4 + CO2 (16) Fe 2 SiOFe2SiO4 4 ` 2CaO + 2CaO ` + 2C 2C = 2Fe +Ca2SiO4 ` Ca 2 SiO + 2CO 4 ` 2CO (17) (17) FeAl 2 O 4 ` CaO ` C Fe ` CaAl 2 O 4 ` CO (18) (18) FeAl2O4 + CaO + C = Fe + CaAl2O4 + CO Equation (1) Equation (4) Fe 3 O 4 CO (%) CO+CO Equation (5) T 1 T 2 Fe 20 FeO Equation (3) 0 Equation (2) -Fe 3 O 4 -FeO -Fe Temperature ( C) 4. Gas phase equilibrium of oxides reduced by solid carbon. 4. Gas-phase equilibrium of oxides reduced by solid carbon.

6 Metals 2016, 6, x 6 of 12 Metals 2016, 6, 57 6 of 12 Metals 2016, 6, x 6 of 12 Δ r G Ө m (kj mol -1 ) Equation (11) Equation (12) (11) 0 Equation (9) Equation (7) Equation Equation (12)(6) Equation 0 (16) Equation Equation (9) (8) Equation Equation (10) (7) Equation (6) Equation (16) Equation (8) - Equation (10) Equation (15) Equation Equation (15) (14) - Equation (14) Equation (18) Equation (18) Equation (13) Equation (13) Δ r G Ө m (kj mol -1 ) o C 5 o C 830 o C 830 o Equation Equation (17) (17) Temperature (( o C) o C) 5. of of (6) (18) by FactSage Calculations of Gibbs energy changes of Equations (6) (18) by FactSage7.0. It can be illustrated that it is difficult to separate aluminum and silica by reducing It can high aluminum be illustrated that hematite limonite it is difficult to ore only separate under aluminum a CO and atmosphere; silica solid carbon and by alkaline earth reducing high aluminum high-aluminum oxide should hematite limonite hematite-limonite be blended to enhance ore only under reduction a CO process atmosphere; to improve solid carbon separation and alkaline earth alkaline-earth of oxide should impurities. be blended to enhance reduction process to improve separation of impurities. 3. Results and Discussion 3. Results and Discussion 3.1. Dosage of Reductant 3.1. Dosage The of Reductant effect of reductant dosage on reduction and separation processes is shown in 6. It shows that optimum metallization degree, grade and and recovery were achieved is shown at 1.0 inof The effect of reductant dosage on reduction and separation processes is shown in 6. It 6. carbon to (C/Fe) ratio, where grade is.58%. However, only part of was shows It shows that that optimum metallization degree, grade and andrecovery were achieved at at1.0 of reduced to metallic ; both metallization degree and recovery were lower than 60%. carbon to (C/Fe) ratio, where grade is is.58%. However, only part of was reduced to metallic ; both metallization degree degree and recovery were lower than 60%. grade of metal powder metallization recoverydegree grade of metal powder recovery Index (%) Index (%) C/Fe ratio 6. Effect of C/Fe ratio on reduction and separation indexes (reducing at 1300 C for 20 min) C/Fe ratio 6. Effect of C/Fe ratio ratioon on reductionand separation indexes (reducing at at 1300 C C for for 20 min).

7 Metals 2016, 6, 57 x 7 of of The coal bearing process is an efficient means to enhance reduction of complex ore because The coal-bearing process can process be self sufficient is an efficient chemically means to to enhance complete reduction task of of metallization complex if ore because required at process is supplied can be[22,23]. self-sufficient Based on chemically rmodynamic to complete analysis, different task of metallization coal bearing if required ratios were heat investigated is supplied [22,23]. to reveal Based effects on rmodynamic of coal bearing analysis, ratio on different reduction coal-bearing and separation ratios were process, investigated where tocoal bearing reveal effects of ratio coal-bearing was defined ratio as on division reduction of andes separation between process, where bearing coal coal-bearing and ore. Results ratio in was defined 7 show as that division metallization of es degree between of reduced bearing pellets coal is and elevated dramatically ore. Results in by coal bearing 7 show that ratio metallization increases from degree 0% 15%, of reduced and pellets recovery is elevated is over dramatically % when byratio coal-bearing is 15%. However, ratio increases grade fromof 0% 15%, metal and powder recovery after magnetic is over separation % when is on ratio a downward is 15%. However, trend when ratio gradeis of beyond metal 15% powder because after of magnetic high impurity separation contents is onin a downward coal ash. trend when ratio is beyond 15% because of high impurity contents in coal ash. Index (%) metallization degree grade of metal powder recovery Coal-bearing ratio (%) 7. Effect of coal bearing ratio on reduction and separation indexes (reducing at 7. Effect of coal-bearing ratio on reduction and separation indexes (reducing at 1300 C for C min for 20 min a total C/Fe a total C/Fe ratio of 1.0). ratio of 1.0). Fayalite and hercynite are detected in reduced pellets of blends out coal bearing, as Fayalite and hercynite are detected in reduced pellets of blends out coal-bearing, illustrated in 8, leading to a lower metallization degree and recovery. Due to low as illustrated in 8, leading to a lower metallization degree and recovery. Due to melting temperature of fayalite, liquid phase formed easily in reduction temperature range, low melting temperature of fayalite, liquid phase formed easily in reduction temperature and micromelting appearance was observed in reduced pellets. Distributions of grains in range, and micromelting appearance was observed in reduced pellets. Distributions of grains 9 demonstrate that metallic grains exist at outer layer of reduced pellets out in coal bearing, 9 demonstrate and inner that layer metallic is filled grains fayalite exist at and outer hercynite. layer In of reduced contrast, pellets dominant out coal-bearing, minerals are reduced and to inner metallic layer is filled at both fayalite inner and and outer hercynite. layer by In blending contrast, 15% dominant coal fines in minerals pellet are feed, reduced and minor to metallic minerals at both combine inner and outer Al2O3 to form layer a by trace blending of hercynite. 15% coal Mullite fines in is pellet main gangue feed, and mineral, minor which minerals inhibits combine grain growth Al 2 O of 3 to metal form a trace to a of certain hercynite. extent, Mullite so that is re main are gangue still some mineral, metallic which grains inhibits fine grain particle growth size of metal less than 20 to μm, a certain resulting extent, in so that low re grade are still of some metal metallic powder grains after magnetic fine particle separation. size less than 20 µm, resulting in low grade of metal powder after magnetic separation.

8 Metals 2016, 6, 57 8 of 12 Metals 2016, 6, x 8 of 12 Metals 2016, 6, 57 8 of 12 Without coal-bearing - Iron - Fayalite Without coal-bearing - - Hercynite Iron - - Mullite Fayalite - Hercynite - Mullite With 15% coal-bearing With 15% coal-bearing Intensity Intensity (Counts) (Counts) CuKα Two-Theta (deg) XRD 8. XRD patterns patterns CuK of reduced of reduced pellets pellets Two-Theta different different (deg) coal-bearing coal bearing ratios ratios (reducing (reducing at 1300 at C 1300 C 8. for XRD 20 min patterns a total of reduced C/Fe pellets ratio of 1.0). different coal bearing ratios (reducing at for 20 min a total C/Fe ratio of 1.0) C for 20 min a total C/Fe ratio of 1.0). 9. Distributions of grains in reduced pellets under different coal bearing ratios ((a,b) inner 9. Distributions and outer layer of out grains coal bearing; reduced (c,d) pellets inner under and outer different layer coal bearing a 15% coal bearing ratios 9. ratio). Distributions of grains in reduced pellets under different coal-bearing ratios ((a,b) inner and outer layer out coal bearing; (c,d) inner and outer layer a 15% coal bearing ((a,b) inner ratio). and outer layer out coal-bearing; (c,d) inner and outer layer a 15% coal-bearing 3.2. Optimization ratio). of Reduction Parameters 3.2. Optimization of Reduction Parameters The grade of metal powder was raised from 81.83% 91.05% by increasing 3.2. Optimization reduction The temperature ofgrade Reduction of from metal Parameters powder C, as demonstrated was raised from in 81.83% 91.05% 10. The reduction by increasing temperature The is reduction an important grade temperature factor of metal that from affects powder reaction C, was as raised demonstrated kinetics; from a higher 81.83% 91.05% in reduction 10. The temperature byreduction increasing can temperature promote reduction is an migration important and factor growth that affects rate of metal reaction kinetics; grains during a higher reduction temperature process. By can considering promote temperature from C, as demonstrated in 10. The reduction temperature is an economic migration efficiency, and growth reduction rate of temperature metal grains was suggested during to be reduction 1350 C process. for furr By research. considering important factor that affects reaction kinetics; a higher reduction temperature can promote economic efficiency, reduction temperature was suggested to be 1350 C for furr research. migration and growth rate of metal grains during reduction process. By considering economic efficiency, reduction temperature was suggested to be 1350 C for furr research.

9 Metals 2016, 6, 57 9 of 12 Metals 2016, 6, x 9 of 12 Metals 2016, 6, x 9 of 12 Index Index (%) (%) metallization degree 75 metallization grade of metal degree 75 powder recovery grade of metal powder 70 recovery Reduction 1300 temperature 1350( o C) 1400 Reduction temperature ( o C) 10. Effect of reduction temperature on reduction and separation indexes (reducing for Effect min 10. Effect of a total of C/Fe reduction reduction temperature ratio temperature on of 1.0, 15% coal on blended reduction reduction in pellets). and and separation separation indexes indexes (reducing (reducing for for 20 min 20 min a total a total C/Fe C/Fe ratio ratio of of 1.0, 1.0, 15% coal blended in inpellets). 11 presents effect of reduction duration on indexes of reduction and separation 11 presents processes, 11 presents which effect shows effect of reduction that of reduction metallization duration duration degree on indexes and indexes of recovery of reduction were reduction and enhanced separation and processes, by separation prolonging processes, which shows reduction which that duration shows that metallization from metallization degree min. and Furr degree extension and recovery of were recovery enhanced reduction were time enhanced by prolonging had a reduction slightly by prolonging negative reduction durationimpact from on duration min. indexes. from FurrTherefore, min. Furr extension of reducing extension reduction at 1350 of time C reduction had for a25 time slightly min had was a negative recommended, slightly negative where impact metal on indexes. powder Therefore, was manufactured 85.00% of grade at a impact on indexes. Therefore, reducing at 1350 C reducing at 1350 C for 25 min was for 25 min was recommended, where metal recovery recommended, of 92.05%, where respectively. metal powder was manufactured 85.00% of grade at a powder recovery was of 92.05%, manufactured respectively % of grade at a recovery of 92.05%, respectively. Index Index (%) (%) metallization degree metallization grade of metal degree powder recovery grade of metal powder recovery Reduction duration 25 (min) 30 Reduction duration (min) 11. Effect of reduction duration on reduction and separation indexes (reducing at 1350 C a total 11. Effect C/Fe of reduction ratio of duration 1.0, 15% coal on blended reduction in pellets). and separation indexes (reducing at 1350 C 11. Effect of reduction duration on reduction and separation indexes (reducing at 1350 C a total C/Fe ratio of 1.0, 15% coal blended in pellets). a total C/Fe ratio of 1.0, 15% coal blended in pellets) Effect of Binary Basicity 3.3. Effect of Binary Basicity 3.3. EffectThe of Binary effects Basicity of binary basicity (ratio of CaO/SiO2) on reduction and separation process are shown The in effects of 12. binary The metallization basicity (ratio degree of CaO/SiO2) of reduced on pellets reduction increases and slightly separation and process n keeps are steady The shown effects when in ofbasicity binary 12. The is basicity metallization elevated (ratio from of degree 0.02 CaO/SiO (natural of reduced 2 ) on basicity) pellets reduction to increases 0.5; and slightly beneficiation separation and n of process keeps are shown concentrate steady in when is 12. improved, The metallization is where elevated from degree 0.02 grade of (natural keeps reduced increasing basicity) pelletsto from increases 0.5; 85.00% 92.64% beneficiation slightly and while of n keeps steady recovery concentrate when basicity of is improved, is isover elevated %. where XRD from of 0.02 (natural reduced grade basicity) keeps pellets increasing to 0.5; 0.3 basicity from beneficiation 85.00% 92.64% in of 13 proves while concentrate that is improved, most recovery where of minerals is over are reduced %. grade XRD keeps to metallic of increasing reduced ; impurities, pellets from 85.00% 92.64% such 0.3 as basicity alumina while and silica, recovery 13 proves react of that is calcium most oxide minerals to form are reduced anorthite to (CaAl2Si2O8); metallic ; and impurities, rest such of as alumina exists and silica, as corundum. react over %. XRD of reduced pellets 0.3 basicity in 13 proves that most minerals Metallic calcium oxide grains to form migrated anorthite and (CaAl2Si2O8); agglomerated and toger rest to of generate alumina coarser exists particles as corundum. are reduced to metallic ; impurities, such as alumina and silica, react calcium oxide (refer to to form Metallic grains migrated and agglomerated toger to generate coarser particles (refer to anorthite (CaAl 2 Si 2 O 8 ); and rest of alumina exists as corundum. Metallic grains migrated and agglomerated toger to generate coarser particles (refer to 14), contributing to higher

10 Metals 2016, 6, of 12 Metals 2016, 6, of 12 Metals 2016, 6, of 12 grade and recovery in magnetic separation because of higher liberation degree of metal 2016, 14),6, 57 contributing to higher grade and recovery in magnetic separation because 10 of of Metals 12 grains in grinding. 14), contributing to higher grade and recovery in magnetic separation because of higher liberation degree of metal grains in grinding. liberation 14), contributing to higher recovery in magnetic separation because of higher degree of metal grade grainsand in grinding. higher liberation degree of metal grains in grinding. Index (%)(%) Index (%) Index metallization degree grade of degree metal powder metallization recovery grade of metal metallization degree powder recovery grade 0.3 of metal 0.4 powder 0.5 recovery Binary basicity Binary basicity 0.0 on and indexes 0.5(reducing at 1350 C for 12. Effect of binary basicity reduction separation 12. Effect of binary basicity on reduction separation indexes (reducing at 1350 C for Binaryand basicity Effect of C/Fe binary basicity 15% reduction and separation indexes (reducing at 1350 C for 25 min 12. a total ratio on of 1.0, coal blended in pellets). 25 min a total C/Fe ratio 15% coalblended blended pellets). 25 min a total ratioof of1.0, 1.0, coal inin pellets). 12. Effect ofc/fe binary basicity on 15% reduction and separation indexes (reducing at 1350 C for 25 min a total C/Fe 0.3 ratio of 1.0,basicity 15% coal blended in pellets). of binary - Iron Intensity (Counts) Intensity (Counts) Intensity (Counts) 0.3 of binary basicity 0.3 of binary basicity - Iron Anorthite -- Anorthite Corundum Iron -- Corundum Anorthite - Corundum CuK Two-Theta (deg) Two-Theta CuK 20 (deg) Main minerals in reduced pellets a binary basicity of 0.3 detected by XRD. CuK Two-Theta (deg) 13. Main minerals in reduced pellets a binary basicity of 0.3 detected by XRD. 13. Main minerals in reduced pellets a binary basicity of 0.3 detected by XRD. 13. Main minerals in reduced pellets a binary basicity of 0.3 detected by XRD. 14. Distributions of metallic grains in reduced pellets at 0.3 binary basicity 14. and Distributions metallic Cgrains in reduced pellets at 0.3ratio binary basicity ((a,b) inner outer layer,of reducing at 1350 for 25 min a total C/Fe of 1.0). ((a,b) outer layer, of reducing at 1350 for 25 min a total C/Fe of 1.0).basicity inner 14. and Distributions metallic Cgrains in reduced pellets at 0.3ratio binary 14. Distributions of metallic grains in reduced pellets at 0.3 binary basicity ((a,b) inner and ((a,b) inner and outer layer, reducing at 1350 C for 25 min a total C/Fe ratio of 1.0). outer layer, reducing at 1350 C for 25 min a total C/Fe ratio of 1.0).

11 Metals 2016, 6, of 12 In summary, optimum conditions were recommended as follows: preparing green balls by blending 15% coal and CaCO 3 at 0.3 basicity ore, reducing dried pellets at 1350 C for 20 min a total C/Fe ratio of 1.0 and magnetic separating of ground reduced pellets a size up to 95% passing at mm and at a 0.10-T magnetic field intensity. Table 3 presents chemical compositions of final products. The metal powder assaying.46% Fe total was achieved at an overall recovery of.25%. The final product can be applied as burden for an electric arc furnace, and nonmagnetic tailings can be used to produce ceramic and glass [24,25]; green utilization of high-aluminum resources is probably achieved. Table 3. Chemical compositions of final products (wt. %). Sample Fe total Al 2 O 3 SiO 2 CaO MgO MnO K 2 O Na 2 O P S Metal powder Nonmagnetic tailings Conclusions (1) A high-aluminum hematite-limonite ore, assaying 41.92% Fe total, 13.74% Al 2 O 3 and 13.96% SiO 2, was used as raw material to manufacture metal powder, where minerals mainly occur in hematite and goethite and aluminum minerals exist in kaolinite. Hematite is closely disseminated kaolinite, leading to harder separation between and alumina. (2) Good quality metal powder assaying.46% Fe total was produced at an overall recovery of.25% under following conditions: balling mixture of ore 15% coal and CaCO 3 fines at 0.3 basicity, reducing dried pellets at 1350 C for 25 min and a total C/Fe ratio of 1.0, grinding reduced pellets up to 95% passing at mm and magnetic separation in a Davis Tube at 0.10 T of magnetic field intensity. (3) The process provides a potential way to utilize high-aluminum ore, where good quality metal powder can be used as burden for EAF and nonmagnetic tailings can be applied in ceramic and glass industries. Acknowledgments: This work was financially supported by Co-Innovation Center for Clean and Efficient Utilization of Strategic Metal Mineral Resources and National Key Technology R&D Program of China (No. 2013BAB03B04). Author Contributions: Xianlin Zhou and Deqing Zhu conceived of and designed experiments. Xianlin Zhou, Yanhong Luo and Xinqi Liu performed experiments. Xianlin Zhou and Jian Pan analyzed data. Deqing Zhu and Jian Pan contributed materials. Xianlin Zhou wrote paper. Conflicts of Interest: The authors declare no conflict of interest. References 1. Steel Statistical Yearbook 2015; World Steel Association: Brussels, Belgium, 2015; p Zhang, K.; Kleit, A.N. Mining rate optimization considering stockpiling: A oretical economics and real option model. Resour. Policy 2016, 47, [CrossRef] 3. Jiang, T.; Liu, M.; Li, G.; Sun, N.; Zeng, J.; Qiu, G. Novel process for treatment of high-aluminum limonite ore by reduction roasting addition of sodium salts. Chin. J. Nonferr. Met. 2010, 20, Shu, C.; Tang, Y.; Wang, Z.; Zhang, Q. Research on sodium salt roasting-acid leaching technology for high-alumina and high-silicon refractory limonite. Min. Metall. Eng. 2012, 32, Lu, L.; Holmes, R.J.; Manuel, J.R. Effects of alumina on sintering performance of hematite ores. ISIJ Int. 2007, 47, [CrossRef] 6. Tian, Q.; Shi, Y. Research on desulfurizing capacity of high Al 2 O 3 content slag. Ironmaking 1998, 17, Raghukumar, C.; Tripathy, S.K.; Mohanan, S. Beneficiation of Indian high alumina ore fines A case study. Int. J. Min. Eng. Miner. Process. 2012, 1, 94. [CrossRef]

12 Metals 2016, 6, of Thella, J.S.; Mukherjee, A.K.; Srikakulapu, N.G. Processing of high alumina ore slimes using classification and flotation. Powder Technol. 2012, 217, [CrossRef] 9. Jain, V.; Rai, B.; Waghmare, U.V.; Tammishetti, V. Pradip Processing of Alumina-Rich Iron Ore Slimes: Is Selective Dispersion-Flocculation-Flotation Solution We Are Looking for Challenging Problem Facing Indian Iron and Steel Industry? Trans. Indian Inst. Met. 2013, 66, [CrossRef] 10. Chun, T.J.; Zhu, D.Q.; Pan, J.; He, Z. Preparation of metallic powder from red mud by sodium salt roasting and magnetic separation. Can. Metall. Quart. 2014, 53, [CrossRef] 11. Li, G.; Zhou, T.; Liu, M.; Jiang, T.; Fan, X. Novel process and mechanisms of aluminum- separation of high-aluminum limonite ore. Chin. J. Nonferr. Met. 2008, 18, Li, G.; Liu, M.; Jiang, T.; Zhou, T.; Fan, X. Mineralogy charateristics and separation of aluminum and of high-aluminum ores. J. Cent. South Univ. 2009, 18, Jiang, T.; Liu, M.; Li, G.; Sun, N.; Zeng, J.; Qiu, G. Effects of sodium-salt on Al-Fe separation by reduction roasting for high-aluminum content limonite. Chin. J. Nonferr. Met. 2010, 20, Li, G.; Jiang, T.; Liu, M.; Zhou, T.; Fan, X.; Qiu, G. Beneficiation of high-aluminium-content hematite ore by soda ash roasting. Min. Proc. Ext. Met. Rev. 2010, 31, [CrossRef] 15. Chun, T.; Long, H.; Li, J. Alumina- separation of high alumina ore by carbormic reduction and magnetic separation. Sep. Sci. Technol. 2015, 50, [CrossRef] 16. Stjernberg, J.; Olivas-Ogaz, M.A.; Antti, M.L.; Ion, J.C.; Lindblom, B. Laboratory scale study of degradation of mullite/corundum refractories by reaction alkali-doped deposit materials. Ceram. Int. 2013, 39, [CrossRef] 17. Stjernberg, J.; Antti, M.; Nordin, L.; Odén, M. Degradation of refractory bricks used as rmal insulation in rotary kilns for ore pellet production. Int. J. Appl. Ceram. Technol. 2009, 6, [CrossRef] 18. Zhu, D.; Zhou, X.; Pan, J.; Luo, Y. Direct reduction and beneficiation of a refractory siderite lump. Miner. Process. Extr. Metall. 2014, 123, [CrossRef] 19. Zhou, X.; Zhu, D.; Pan, J.; Wu, T. Utilization of waste copper slag to produce directly reduced for wearing resistant steel. ISIJ Int. 2015, 55, Huang, X. Ferrous Metallurgy Theory, 4th ed.; Metallurgical Industry Press: Beijing, China, 2005; pp Zhang, B.; Wang, Z.; Gong, X.; Guo, Z. Study on reduction of high-aluminum ores gas in a fluidized bed (in Chinese). J. Chin. Soc. Rare Earth. 2013, 30, Wei, Y.; Sun, T.; Kou, J.; Yu, W.; Cao, Y. Effect of coal dosage on direct reduction roasting of refractory ore briquettes. J. Cent. South Univ. 2013, 44, Lu, W.; Huang, D.F. The evolution of making process based on coal-containing ore agglomerates. ISIJ Int. 2001, 41, [CrossRef] 24. Gu, X.; Dong, W.; Chen, Y.; Li, Y. Investigation on glass-anorthite insulating composites low dielectric constant and low sintering temperature. Rare Metal Mater. Eng. 2007, 36, Liu, Q.; Pan, Z.; Li, Q.; Ruan, Y. Preparation of anorthite lightweight rmal insulating brick and formation process of anorthite. Bull. Chin. Ceram. Soc. 2010, 29, by authors; licensee MDPI, Basel, Switzerland. This article is an open access article distributed under terms and conditions of Creative Commons by Attribution (CC-BY) license (