An Innovative Technique for Comprehensive Utilization of High Aluminum Iron Ore Via Pre-Reduced-Smelting Separation-Alkaline

metals

Article An Innovative Technique for Comprehensive Utilization of High Aluminum Iron Ore via Pre-Reduced-Smelting Separation-Alkaline Leaching Process: Part I: Pre-Reduced-Smelting Separation to Recover Iron

Siwei Li, Jian Pan , Deqing Zhu *, Zhengqi Guo *, Yue Shi, Jianlei Chou and Jiwei Xu

School of Minerals Processing and Bioengineering, Central South University, Changsha 410083, China; [email protected] (S.L.); [email protected] (J.P.); [email protected] (Y.S.); [email protected] (J.C.); [email protected] (J.X.) * Correspondence: [email protected] (D.Z.); [email protected] (Z.G.); Tel.: +86-135-0848-0691(D.Z.); +86-185-7313-1417 (Z.G.) Received: 22 November 2019; Accepted: 24 December 2019; Published: 28 December 2019

Abstract: In this study, a novel process was established for extraction of Fe and Al from a complex high aluminum iron ore (33.43% Fetotal and 19.09% Al2O3). The main steps in the proposed process included pre-reducing high alumina iron ore and subsequent smelting to produce pig iron and rich-alumina slag, followed by alkaline leaching of the slag to obtain sodium aluminate solution and a clean slag. When smelting the pre-reduced high alumina iron ore pellets at 1625 ◦C for 30 min with a slag basicity of 0.40, the pig iron yielded 97.08% Fe and extracted 0.13% Al2O3, together with an iron recovery of 94.54%. In addition, more than 68.93% Al2O3 was recovered by leaching the slag, which was achieved by ﬁrstly roasted the slag at 900 ◦C for 2 h and then alkaline leaching at 95 ◦C for 2 h with a liquid-to-solid ratio of 10 mL/g. In addition, the alkaline leaching slag could potentially be used as raw material for construction purpose, which mainly consisted of SiO2 and CaO.

Keywords: high-aluminum iron ore; smelting separation; modiﬁcation; alkaline leaching

1. Introduction With the rapid increasing in demand for metals, the global reserves of high-grade ores are shrinking at an alarming rate. Therefore, it is urgent to develop a novel process to upgrade the complex iron resources [1,2]. High aluminum iron ore (HAIO), a kind of polymetallic complex resources, is widely distributed in nature, which has a high comprehensive utilization value. Moreover, it is abundantly found in China, Australian, India and Indonesia [3]. Typically, it contains 5–20% Al2O3 and 30–60% Fe, indicating that it is difficult to be used as a raw material for ironmaking [4,5]. Thus, it is important to remove alumina from HAIO. The recovery of valuable metals, such as Fe and Al, from HAIO has attracted increasing attention. Many technologies have been reported in previous literature, including physical beneficiation, biological leaching, and smelting. The beneficiation methods include both physical [6–8] and chemical [9] processes. Thella et al. [7] adopted flotation process to treat HAIO slime (4.0% Al2O3 and 61.99% Fe), and the findings demonstrated that the iron recovery is only 70% and the content of Al2O3 in final concentrate remained as high as 2.66%. Generally, it is difficult to effectively remove aluminum by the using the conventional beneficiation process, due to the dissemination of aluminum minerals is complex and micro-grained [10]. Pradhan et al. [2] used Bacillus circulars and Aspergillus niger to

Metals 2020, 10, 57; doi:10.3390/met10010057 www.mdpi.com/journal/metals Metals 2020, 10, 57 2 of 17

remove aluminum from iron ore containing Fetotal 52.94%, 9.95% Al2O3 and 6.1% SiO2, and only 40% alumina was removed after 6 or 15 days of in situ leaching at 10% pulp density. Biological method could play a certain role in the separation of iron from alumina, but its effect may be poorer, especially the period is too lengthy compared to smelting processes. Smelting processes consist of coal-based direct reduction-magnetic separation, roasting-magnetic separation, and smelting separation [11–13]. It is effective to recover iron from low-grade refractory iron ore through coal-based direct reduction-magnetic separation. However, there is a problem that exists in this process, low reduction rate [14,15]. Zhou et al. [13] proposed a high temperature reduction followed by magnetic separation for high aluminum hematite-limonite ore treatment. Their findings indicated that approximately 90.25% of iron recovery was obtained, but the recovery of Al2O3 was not considered. Sellaeg et al. [16] adopted a direct smelting reduction process to recover iron from red mud, and the result showed that only 70% of iron was recovered and an Al2O3-containing slag was obtained. He et al. [17] proposed a slag-iron bath smelting reduction process for the treatment of alumina-rich iron ore, and more than 90% iron was recovered using this method. It is worth noting that the smelting separation process can achieve the separation of iron and alumina. However, the slag could not be effectively used, resulting in a difficulty of extracting aluminum. Considering that the existing processes for HAIO treatment have highly energy consumption and are economically impractical, it is urgent to develop a more economical and valid process. In this work, a new process was developed through the pre-reduction of high alumina iron ore and subsequent smelting to produce pig iron and rich-alumina slag, followed by alkali leaching to extract aluminum and generate a clean slag for cement industry.

2. Materials and Methods

2.1. Materials

2.1.1. High Aluminum Iron Ore The HAIO used in this study was supplied by Guang Xi Province, China. To remove all free moisture, the sample was pre-dried for 10 h at 100 ◦C prior to chemical analysis, and the ﬁndings are shown in Table1. The major ingredients of HAIO included 33.43 wt. % Fe, 19.09 wt. % Al 2O3 and 18.36% SiO2.

Table 1. Chemical compositions of high aluminum iron ore (HAIO) (wt. %).

Element TFe Al2O3 SiO2 K2O Na2O MgO CaO P S LOI Content 33.43 19.09 18.36 0.12 0.033 0.063 0.039 0.20 0.028 12.34 LOI: Loss on ignition.

As shown in Figure1, the minerals compositions of the main phases of the iron ore cover goethite (35%), hematite (33%) and kaolinite (32%). The distributions of iron and aluminum in the ore are summarized in Tables2 and3, showing that the iron is predominantly found in the form of hematite and goethite, while aluminum is dominantly presented in the form of kaolinite. Figure2 reveals the microstructure of HAIO. The results of EDS analysis for the areas in Figure2 is demonstrated in Table4. It was observed that the main iron minerals existed within 10 µm and embedded closely with gangue minerals, which in turn could impart a great challenge for the removal of aluminum from HAIO through physical beneﬁciation processes. 3 of 17 3 of 17

Minerals Diasporite Gibbsite Kaolinite TAl2O3 Al MineralsContent Diasporite3.03 Gibbsite2.59 Kaolinite16.16 21.78TAl2 O3 FractionAl Content 13.933.03 11.892.59 74.1816.16 10021.78 Fraction 13.93 11.89 74.18 100 Metals 2020, 10, 57 3 of 17

Figure 1. XRD diffraction diﬀraction pattern of HAIO.HAIO. Figure 1. XRD diffraction pattern of HAIO.

Figure 2. MMicrostructureicrostructure of the HAIO (G-Goethite,(G-Goethite, H-Hematite,H-Hematite, K-Kaolinite).K-Kaolinite). Figure 2. Microstructure of the HAIO (G-Goethite, H-Hematite, K-Kaolinite). Table 2.TableDistributions 4. The EDS of ironanalysis in mineral results phasesfor areas of thein Figure HAIO 2 (wt.. %). Table 4. The EDS analysis results for areas in Figure 2. AreaMinerals No. Iron CarbonateElemental IronCompositions/ SulﬁdeMagnetite (atomic conc, Hematite %) /GoethiteMineral Fayalite Phase TFe FeArea Content No. 0.14Elemental C 0.020ompositions/ 0.050(atomic conc, %) 37.46Mineral 0.47 Phase 38.14 Fraction 0.36 0.06 0.14 98.21 1.23 100

Metals 2020, 10, 57 4 of 17

Table 3. Distributions of aluminum in mineral phases of the HAIO (wt. %).

Minerals Diasporite Gibbsite Kaolinite TAl2O3 Al Content 3.03 2.59 16.16 21.78 Fraction 13.93 11.89 74.18 100

Table 4. The EDS analysis results for areas in Figure2.

Elemental Compositions/(atomic conc, %) Area No. Mineral Phase Fe Al Si O 1 33.49 6.57 4.89 55.13 hematite 2 18.19 12.51 8.66 60.64 goethite 3 5.34 18.53 18.19 57.94 kaolinite 4 6.82 18.34 18.34 56.50 kaolinite 5 18.82 14.66 12.18 54.34 hematite/kaolinite 6 30.77 8.34 6.69 54.20 hematite 7 29.29 8.47 6.50 55.74 hematite 8 20.21 12.58 6.42 60.79 goethite 9 38.55 3.18 57.89 0.38 hematite 10 4.66 19.31 19.70 56.32 kaolinite 11 18.19 12.51 8.66 60.64 goethite 12 5.34 18.53 18.19 57.94 kaolinite 13 20.21 12.58 6.42 60.79 goethite 14 32.96 7.73 52.36 6.95 hematite

2.1.2. Flux and Additive In this study, limestone was employed as a ﬂux to modify the binary basicity of pre-reduced-smelting separation process, which contains 55.73% of CaO and 41.52% of loss on ignition. Analytical grade Na2CO3 was used as the additive in this study for reconstructing the mineral phases of smelting separation slag by modifying the ﬁring process.

2.1.3. Reductant Soft coal was selected as a reductant in this study, which was crushed and screened to a size of 1–5 mm. A proximate analyzer (SDTGA5000, Hunan Sundy, Changsha, China) was used for the proximate analysis of soft coal, while the chemical compositions of ash were evaluated by chemical titration method [18,19]. Table5 summarizes the results of proximate analysis and the major chemical compositions of ash.

Table 5. Proximate analysis of soft coal and its ash chemical compositions (wt. %).

Proximate Analysis * Main Chemical Compositions of ash

FCad Mad Ad Vdaf Fe2O3 SiO2 Al2O3 CaO 52.12 12.98 4.49 30.41 16.86 40.19 15.15 26.55 * FCad: Fixed Carbon content; Mad: moisture; Ad: ash-dry content; Vdaf: volatile content.

2.2. Experimental Methods A schematic ﬂow diagram for the treatment process of the proposed HAIO is illustrated in Figure3, which was established based on the following steps: (1) pre-reduction of the HAIO pellets followed by smelting separation to recover Fe; (2) modiﬁcation the smelting slag-alkaline leaching to extract Al. Metals 2020, 10, 57 5 of 17 5 of 17

FigureFigure 3. Experiment 3. Experiment flow sheet flow sheetof the ofproposed the proposed HAIO HAIO treatment treatment process process.. 2.2.1. Pre-Reduction-Smelting Separation Process 2.2.1. Pre-Reduction-Smelting Separation Process HAIO was sufficiently mixed with a ratio of flux. The mixture was pelletized into 12–16 mm green HAIO was sufficiently mixed with a ratio of flux. The mixture was pelletized into 12–16 mm balls using a disc pelletizer (diameter: 800 mm and depth: 200 mm), and then dried in a vacuum oven green balls using a disc pelletizer (diameter: 800 mm and depth: 200 mm), and then dried in a vacuum at 110 C for 4 h until their weight remained constant. The dried pellets (50 g) with 75 g coals were oven at 110◦ °C for 4 h until their weight remained constant. The dried pellets (50 g) with 75 g coals loaded into a corundum crucible (diameter: 80 mm and height: 80 mm) and roasted in a muffle furnace were loaded into a corundum crucible (diameter: 80 mm and height: 80 mm) and roasted in a muffle (model: KSY-12-18, Hefei Kejing Material Technology Co., Ltd., Hefei, China) at pre-set temperature furnace (model: KSY-12-18, Hefei Kejing Material Technology Co., Ltd., Hefei, China) at pre-set (800–1100 C) for a specific period of time (15–120 min) [20]. Afterwards, the reduced pellets were temperature (800◦ –1100 °C ) for a specific period of time (15–120 min) [20]. Afterwards, the reduced cooled down to room temperature under nitrogen atmosphere, and then prepared for the subsequent pellets were cooled down to room temperature under nitrogen atmosphere, and then prepared for smelting separation process. The metallization rates of iron were determined as the following: the subsequent smelting separation process. The metallization rates of iron were determined as the following: MFe ηFe = 100% (1) 푀퐹푒 TFe × 휂 = × 100% (1) 퐹푒 푇퐹푒 where ηFe is the metallization rate of iron; TFe is the content of total iron in the pre-reduced, which was wheredetermined ηFe is the metallization as the equation: rate w(TFe) of iron=; w(MFe)TFe is the+ w(Fecontent content of total in “FeO”);iron in MFethe preis the-reduced, content which of metallic was determinediron in the pre-reduced as the equation: pellets. w(TFe) = w(MFe) + w(Fe content in “FeO”); MFe is the content of metallic ironDue in to the the pre lower-reduced metallization pellets. rates of pre-reduced pellets and the main iron-containing phases Dueare fayalite to the lower and hercynite, metallization a massive rates amountof pre-reduced of reductant pellets should and the be addedmain iron into-containing the smelting phas separationes are fayaliteprocess, and which hercynite is favorable, a massive for the separation amount of of reductant iron and aluminum. should be The added pre-reduced into the pellets smelting (~100 g) separationwith 5–20 process wt. %, which of coals is werefavorable loaded for into the a separation corundum crucibleof iron and (diameter: aluminum. 80 mm The and pre height:-reduced 90 mm), pelletsand (~100 then g) transferred with 5–20 wt. into % a of smelting coals were furnace loaded (model into KSl-1700X-A4,a corundum crucible Hefei Kejing(diameter Material: 80 mm Technology and height:Co., 90 Ltd., mm) Hefei,, and China) then transferred at pre-set temperature into a smelting (1550–1625 furnace◦C) (model for a definite KSl-1700X period-A4, of Hefei time (10–40 Kejing min). MaterialAfter Technology smelting separation, Co., Ltd. the, Hefei, pig iron China and) slagat pre were-set cooled temperature downto (1550 room–1625 temperature °C ) for undera definite nitrogen periodatmosphere. of time (10 The–40 smeltingmin). After separation smelting indexes separation, include the the pig recovery iron and rates slag of ironwere and cooled the content down ofto total roomiron temperature in the pig under iron. Thenitrogen recovery atmosphere rates of. iron The weresmelting determined separation by indexes the following include formula: the recovery rates of iron and the content of total iron in the pig iron. The recovery rates of iron were determined by the following formula: TFe0 y εFe = × 100% (2) TFe × 푇퐹푒0 × 푦 ε = × 100% (2) 퐹푒 푇퐹푒

Metals 2020, 10, 57 6 of 17

where εFe is the recovery rate of iron; TFe0 is the content of total iron in pig iron; y is the yield of pig iron; TFe is the content of total iron in the pre-reduced pellets

2.2.2. The Slag Modification-Alkaline Leaching The main compositions of the pre-reduced pellets are hercynite, fayalite, quartz, and gehlenite, which can be used to reconstruct the mineral phases by adding Na2CO3 and limestone. The slag from the smelting separation process was first ground below 0.074 mm, and then sufficiently mixed with − specific amounts of additives (10 wt. % limestone and 20 wt. % Na2CO3). The mixture containing 5% moisture was prepared in the form of briquettes (diameter: 10mm and height: 10mm), and then dried in a vacuum oven at 110 ◦C for 4 h. After drying, the dried briquettes (30 g) were charged into a corundum crucible (length: 30 mm and height: 20 mm), and then fired in a pipe furnace (model: KSY-12-18, Hefei Kejing Material Technology Co., Ltd., Hefei, China) at 900 ◦C for 120 min for the purpose of modification. Subsequently, the modified slag was cooled down to room temperature. The alkaline leaching tests were performed by mixing the modified slag with 2 mol/L alkaline solution at a liquid-to-solid (L/S) ratio of 10 mL/g and a rotation speed of 300 rpm. The mixture was heated at 95 ◦C in a thermostatic water bath (HH-12468, Changzhou Yuexin Instrument Manufacturing Co., Ltd., Changzhou, China) for 2 h. After filtration, the leach residues were rinsed with deionized water. The recovery rates of Al2O3 were calculated by the following: " # y w ε = 1 1 Al2O3 100 × 100% (3) − w0 ×

ε where Al2O3 is the recovery rate of Al2O3; y1 is the yield of leaching residue; w0 and w1 are the content of Al2O3 in the modiﬁed slag and leaching residue, respectively.

2.3. Characterization of Raw Materials and Products To determine the chemical compositions of HAIO, X-ray ﬂuorescence (XRF, PANalytical Axios; RIGAKU ZSX Priums, PANalytical B.V., Almelo, The Netherlands) was carried out. Mineral compositions and distributions of aluminum and iron in HAIO at various phases were evaluated by chemical phase analysis as described previously [21,22]. The concentrations of metallic iron were determined by chemical titration method according to the national standards (GB/T 24194-2009) [23]. The microstructures of HAIO and smelting separation slag were examined using Leica DMLP optical microscopy, FEI Quata-200 (Leica, Weztlar, Germany) scanning electron microscope (SEM), energy-dispersive spectrometry (EDS; XMAX20, FEI, Eindhoven, The Netherlands), EDAX32 genesis spectrometer (FEI, Eindhoven, The Netherlands) and TESCAN MIRA-3, LMH SEM(FEI, Eindhoven, The Netherlands) The mineral phases were measured by X-ray powder diﬀraction (XRD, RIGAKU, D/Max-2500, Bruker corporation, Madison, WI, USA) with a 2θ scan range from 10◦–70◦, operating at 40 KV and 40 mA with Cu Kα radiation (λ = 0.15418 nm) at 25 ◦C The contents of main mineral of raw material, pre-reduced pellets and slag were calculated by the following formula [24]:

I1 1 xi = xs (4) × IS × k1 where xi and xs are the content of determined mineral and reference substance, respectively; I1 and IS are the diﬀraction peak intensity of determined mineral and reference substance, respectively; k1 is the reference intensity (diﬀraction peak intensity is the result of the collective action of all the atoms in the unit cell). 7 of 17

As shown in Figure 4., the metallization rates of iron were elevated from 8.18% to 40.27% with increasingMetals 2020 , reduction10, 57 temperature (800 °C–1000 °C). Meanwhile, the content of iron was changed7 of 17 slightly. At an elevated temperature, the metallization rate of iron was reduced. This is probably due to the fact that rising temperature is beneficial to enhance the reduction of fayalite (Fe2SiO4) and 3. Results and Discussion hercynite (FeAl2O4), but more liquid phase would be formed if the reduction temperature is too high, which3.1. Recover may Ironbe detrimental from HAIO byfor Pre-Reduction-Smelting diffusion of the reduction Separation gas Process [13]. Therefore, the reduction temperature of 1000 °C was selected for subsequent experiments. 3.1.1.Fig Pre-Reductionure 5 reveals the of Higheffect Aluminas of reduction Iron Oreduration Pellets on the iron grade and metallization rate of the pre-reducedAs shown pellets. in Figure The metallization4, the metallization rates were rates increased of iron were by elevated prolonging from the 8.18% reducing to 40.27% duration with from 15 to 60 min. Further extending the reduction duration could slightly affect the two indices, increasing reduction temperature (800 ◦C–1000 ◦C). Meanwhile, the content of iron was changed indicatingslightly. Atthat an the elevated reductant temperature, is nearly exhausted the metallization and the reduction rate of iron atmosphere was reduced. is diminished This is probably due to a lengthy period of reduction. Based on these finding, a reduction time of 60 min was selected. As a due to the fact that rising temperature is beneﬁcial to enhance the reduction of fayalite (Fe2SiO4) and result, the pre-reduced pellets were manufactured with 39.93% of iron grade and 40.27% of hercynite (FeAl2O4), but more liquid phase would be formed if the reduction temperature is too high, metallizationwhich may be rate. detrimental for diﬀusion of the reduction gas [13]. Therefore, the reduction temperature of 1000 ◦C was selected for subsequent experiments.

Figure 4. Effects of reduction temperature on the pre-reduction indices. (Reducing for 60 min, with a FigureC/Fe mass4. Effect ratios of of reduction 1.50, natural temperature basicity of on 0.0020). the pre-reduction indices. (Reducing for 60 min, with a C/Fe mass ratio of 1.50, natural basicity of 0.0020). Figure5 reveals the e ffects of reduction duration on the iron grade and metallization rate of the pre-reduced pellets. The metallization rates were increased by prolonging the reducing duration from 15 to 60 min. Further extending the reduction duration could slightly affect the two indices, indicating that the reductant is nearly exhausted and the reduction atmosphere is diminished due to a lengthy period of reduction. Based on these finding, a reduction time of 60 min was selected. As a result, the pre-reduced pellets were manufactured with 39.93% of iron grade and 40.27% of metallization rate. Figure6 shows the e ffects of basicity on the iron grade and metallization rate of the pre-reduced pellets. The metallization rates were elevated from 40.99% to 52.42% when the basicity levels were increased from 0.002 (natural basicity) to 0.8. The XRD patterns of pre-reduced pellets at different binary basicity (w(CaO)/w(SiO2)) levels are presented in Figure7. The mineral compositions of the pre-reduced pellets of natural basicity are element iron, fayalite, hercynite, and quartz. The contents of fayalite and hercynite decreased with increasing basicity levels from 0.002 to 0.8. On the contrary, the contents of gehlenite increased with increasing basicity levels (as shown in Table6). These results provide compelling evidence that CaO could enhance the decomposition of fayalite and hercynite, which in turn leads to the improved activity of FeO. And ultimately increased the rates of metallization.

Figure 5. Effect of reduction duration on the pre-reduction indices. (Reduction at 1000 °C, with a C/Fe mass ratio of 1.50, natural basicity of 0.0020).

7 of 17

As shown in Figure 4., the metallization rates of iron were elevated from 8.18% to 40.27% with increasing reduction temperature (800 °C–1000 °C). Meanwhile, the content of iron was changed slightly. At an elevated temperature, the metallization rate of iron was reduced. This is probably due to the fact that rising temperature is beneficial to enhance the reduction of fayalite (Fe2SiO4) and hercynite (FeAl2O4), but more liquid phase would be formed if the reduction temperature is too high, which may be detrimental for diffusion of the reduction gas [13]. Therefore, the reduction temperature of 1000 °C was selected for subsequent experiments. Figure 5 reveals the effects of reduction duration on the iron grade and metallization rate of the pre-reduced pellets. The metallization rates were increased by prolonging the reducing duration from 15 to 60 min. Further extending the reduction duration could slightly affect the two indices, indicating that the reductant is nearly exhausted and the reduction atmosphere is diminished due to a lengthy period of reduction. Based on these finding, a reduction time of 60 min was selected. As a result, the pre-reduced pellets were manufactured with 39.93% of iron grade and 40.27% of metallization rate.

Figure 4. Effects of reduction temperature on the pre-reduction indices. (Reducing for 60 min, with a C/Fe mass ratio of 1.50, natural basicity of 0.0020). Metals 2020, 10, 57 8 of 17

8 of 17

Figure 6 shows the effects of basicity on the iron grade and metallization rate of the pre-reduced pellets . The metallization rates were elevated from 40.99% to 52.42% when the basicity levels were8 of 17 increased from 0.002 (natural basicity) to 0.8. The XRD patterns of pre-reduced pellets at different Figure 6 shows the effects of basicity on the iron grade and metallization rate of the pre-reduced binary basicity (w(CaO)/w(SiO2)) levels are presented in Figure 7. The mineral compositions of the pellets. The metallization rates were elevated from 40.99% to 52.42% when the basicity levels were pre-reduced pellets of natural basicity are element iron, fayalite, hercynite, and quartz. The contents increased from 0.002 (natural basicity) to 0.8. The XRD patterns of pre-reduced pellets at different of fayalite and hercynite decreased with increasing basicity levels from 0.002 to 0.8. On the contrary, binary basicity (w(CaO)/w(SiO2)) levels are presented in Figure 7. The mineral compositions of the the contents of gehlenite increased with increasing basicity levels (as shown in Table 6). These results pre-reduced pellets of natural basicity are element iron, fayalite, hercynite, and quartz. The contents provide compelling evidence that CaO could enhance the decomposition of fayalite and hercynite, of fayalite and hercynite decreased with increasing basicity levels from 0.002 to 0.8. On the contrary, which in turn leads to the improved activity of FeO. And ultimately increased the rates of the contents of gehlenite increased with increasing basicity levels (as shown in Table 6). These results metallization. provide compelling evidence that CaO could enhance the decomposition of fayalite and hercynite, which in turn leads to the improved activity of FeO. And ultimately increased the rates of metallization.

Figure 5. Eﬀect of reduction duration on the pre-reduction indices. (Reduction at 1000 ◦C, with a C/Fe Figuremass ratio5. Effect of 1.50, of reduction natural basicity duration of o 0.0020).n the pre-reduction indices. (Reduction at 1000 °C, with a C/Fe mass ratio of 1.50, natural basicity of 0.0020).

Figure 6. Effects of binary basicity (w(CaO)/w(SiO2)) on the pre-reduction indices (Reduction at 1000 °C for 60 min, with a C/Fe mass ratio of 1.50). Figure 6. Eﬀects of binary basicity (w(CaO)/w(SiO2)) on the pre-reduction indices (Reduction at 1000 ◦C Figurefor 60 6. min, Effect withs of a binary C/Fe mass basicity ratio (w(CaO)/w(SiO of 1.50). 2)) on the pre-reduction indices (Reduction at 1000 °C for 60 min, with a C/Fe mass ratio of 1.50).

Figure 7. XRD pattern of the pre-reduced pellets at diﬀerent binary basicity (w(CaO)/w(SiO2)) (A: Albite Figure 7. XRD pattern of the pre-reduced pellets at different binary basicity (w(CaO)/w(SiO2)) (A: (Na,Ca)Al(Si,Al)3O8; Q: Quartz SiO2; G: Gehlenite Ca2Al2SiO7;H: Hercynite FeAl2O4; F: Fayalite Albite (Na,Ca)Al(Si,Al)3O8; Q: Quartz SiO2; G: Gehlenite Ca2Al2SiO7;H: Hercynite FeAl2O4; F: Fayalite Fe2SiO4; I: element Iron;). Fe2SiO4; I: element Iron;). Figure 7. XRD pattern of the pre-reduced pellets at different binary basicity (w(CaO)/w(SiO2)) (A: Albite (Na,Ca)Al(Si,Al)Table 6.3O The8; Q: contentQuartz ofSiO main2; G: mineralGehlenites in Ca the2Al Fig2SiOure7;H: 7 (wt.Hercynite %). FeAl2O4; F: Fayalite Fe2SiO4; I: element Iron;). Basicity Iron Hercynite Fayalite Quartz Gehlenite Albite Table 6. The content of main minerals in the Figure 7 (wt. %).

Basicity Iron Hercynite Fayalite Quartz Gehlenite Albite

Metals 2020, 10, 57 9 of 17

Table 6. The content of main minerals in the Figure7 (wt. %).

Basicity Iron Hercynite Fayalite Quartz Gehlenite Albite 0.0020 12 58 12 18 - - 0.20 15 45 10 14 8 8 0.40 18 31 9 7 23 12 0.60 13 27 10 8 35 7 0.80 13 19 8 6 47 7

3.1.2. Smelting Separation Process

Thermodynamic Analysis As shown in Figure7, the main iron minerals in the pre-reduced pellets are fayalite and hercynite. Equations (5–12) reveals the potential reactions during the smelting process of pre-reduced pellets. FactSage7.0 (Thermfact/CRCT, Montreal, QC, Canda; GTT-Technologies, Herzogenrath, Germany) was adopted to estimate the changes in the Gibbs free energy of the reactions; and the ﬁndings are presented in Figure8. Based on the thermodynamic calculation, both fayalite and hercynite were diﬃcult to be reduced by CO thermodynamically (Equations (7) and (11)), indicating that they are not able to be reduced in pre-reduction process. However, lower CO concentration might be required to reduce fayalite and hercynite when CaO is existed. In the smelting separation process, solid carbon was employed as the reductant, and thus both fayalite and hercynite could be reduced. Additionally, CaO could promote the reduction of fayalite and hercynite by chemically reacting with Al2O3 and SiO2 to form calcium aluminate (CaAl2O4) and larnite (Ca2SiO4), respectively.

Fe2SiO4 + 2C = 2CO +2Fe + SiO2 (5)

Fe2SiO4 + 2C + 2CaO = 2Fe + 2CO + Ca2SiO4 (6)

1/2Fe2SiO4 + CO = CO2 + Fe + 1/2SiO2 (7)

1/2Fe2SiO4 + CO + CaO = CO2 + Fe + 1/2Ca2SiO4 (8)

FeAl2O4 + C = Fe + Al2O3 + CO (9)

FeAl2O4 + C + CaO = Fe + CO + CaAl2O4 (10)

FeAl2O4 + CO = CO2 + Fe + Al2O3 (11)

FeAl2O4 + CO + CaO = Fe + CO2 + CaAl2O4 (12)

C + CO2 = 2CO (13) According to the chemical composition of pre-reduced pellets, the liquid phase quantity and viscosity at different temperature were calculated by FactSage7.2 (Thermfact/CRCT, Montreal, QC, Canda; GTT-Technologies, Herzogenrath, Germany), and the results are shown in Figure9. As can be seen from that the figure, the liquid phase quantity increased from 85.53% to 100% with elevating smelting temperature from 1400 ◦C to 1550 ◦C under natural basicity. Meanwhile, the viscosity decreased from 2.56 Pa s to 0.92 Pa s. The smelting separation process required the slag to have · · lower viscosity, and its value for ferroalloy production is generally between 0.20 Pa s and 0.50 Pa s. · · When the viscosity of the slag is higher than 1 Pa s, it is hard to separate the iron from the slag [25]. · Nevertheless, the viscosity of the slag decreased from 0.92 Pa s to 0.28 Pa s when the binary basicity · · (w(CaO)/w(SiO2)) level of the slag was increased from natural basicity to 0.80, indicating that addition of CaO can improve the fluidity and viscosity of the slag. 9 of 17

0.0020 12 58 12 18 - - 0.20 15 45 10 14 8 8 0.40 18 31 9 7 23 12 0.60 13 27 10 8 35 7 0.80 13 19 8 6 47 7

3.1.2. Smelting Separation Process

Thermodynamic Analysis As shown in Figure 7, the main iron minerals in the pre-reduced pellets are fayalite and hercynite. Equations (5–12) reveals the potential reactions during the smelting process of pre-reduced pellets. FactSage7.0 (Thermfact/CRCT, Montreal, QC, Canda; GTT-Technologies, Herzogenrath, Germany) was adopted to estimate the changes in the Gibbs free energy of the reactions; and the findings are presented in Figure 8. Based on the thermodynamic calculation, both fayalite and hercynite were difficult to be reduced by CO thermodynamically (Equations (7) and (11)), indicating that they are not able to be reduced in pre-reduction process. However, lower CO concentration might be required to reduce fayalite and hercynite when CaO is existed. In the smelting separation process, solid carbon was employed as the reductant, and thus both fayalite and hercynite could be reduced. Additionally, CaO could promote the reduction of fayalite and hercynite by chemically reacting with Al2O3 and SiO2 to form calcium aluminate (CaAl2O4) and larnite (Ca2SiO4), respectively.

Fe2SiO4 + 2C = 2CO +2Fe + SiO2 (5)

Fe2SiO4 + 2C + 2CaO = 2Fe + 2CO + Ca2SiO4 (6)

1/2Fe2SiO4 + CO = CO2 + Fe + 1/2SiO2 (7)

1/2Fe2SiO4 + CO + CaO = CO2 + Fe + 1/2Ca2SiO4 (8)

FeAl2O4 + C = Fe + Al2O3 + CO (9)

FeAl2O4 + C + CaO = Fe + CO + CaAl2O4 (10)

FeAl2O4 + CO = CO2 + Fe + Al2O3 (11)

FeAl2O4 + CO + CaO = Fe + CO2 + CaAl2O4 (12)

Metals 2020, 10, 57 C + CO2 = 2CO 10(13 of) 17

（a）（b）

θ Figure 8. The correlation of the standard free energy (∆rGm ) with reaction temperature for Equations Figure(5–12). 8. (Thea): Thecorrelation standard of the free standard energy versusfree energy reaction (ΔrG temperature;mθ) with reaction (b): temperature Gas-phase equilibriumfor Equations of (5Equations–12). (a): (7,8,11,12) The standard reduced free by energy CO). versus reaction temperature; (b): Gas-phase equilibrium of Equations (7,8,11,12) reduced by CO).

Figure 9. Effects of temperature on the mass fraction of liquid phase and viscosity. (a)Effects of temperature on the liquid quantity; (b)Effects of temperature on the viscosity (liquid phase quantity represents the mass fraction of liquid phase).

Smelting Separation of the Pre-Reduced Pellets Figure 10 shows the effect of the process parameters on the smelting separation indices. It was found that the grades of pig iron changed slightly with elevating smelting temperature from 1550 ◦C to 1625 ◦C However, the recovery rates of iron increased from 49.83% to 65.40% (as shown in Figure 10a). Compared to the viscosity of 0.92 Pa s at 1550 C, it was only 0.47 Pa s at 1625 C Indeed, the lower · ◦ · ◦ viscosity at higher temperature is favorable for the separation of iron and slag, thus the optimum smelting temperature of 1625 ◦C was selected. Figure 10b reveals the effects of smelting duration on the indices. Notably, the recovery rates of iron increased from 55.39% to 73.06% by prolonging the smelting duration from 10 min to 30 min. With a further extension of the time, the recovery rates were increased slightly. In addition, the iron grade changed slightly throughout the smelting duration. Based on the above results, smelting at 1625 ◦C for 30 min with natural basicity (0.002) was selected. As a result, the pig iron could be generated from 96.53% of iron grade and 73.06% of iron recovery rate. The effects of reductant dosage on the smelting separation process indices are depicted in Figure 11. The grades of the pig iron changed slightly with increasing doses (5–15%) of the reductant. However, the iron recovery rates of pig iron were obviously increased from 62.44% to 78.47%. Further increasing the reductant doses could negatively affect the values of the indices, probably due to fact that the excessive reductant may cause serious carburizing and lower basicity levels of the slag, leading to the

Coatings 2020, 10, x; doi: FOR PEER REVIEW www.mdpi.com/journal/coatings 11 of 17

Metals 2020, 10, 57 11 of 17

decreased iron grade and higher viscosity of the slag. Considering these indices, the optimal reductant 11 of 17 dosage was recommended as 15%.

Figure 10. Effects of smelting separation parameters on the smelting separation of pre-reduced pellets. (a) Smelting at various temperature for 20 min with 10% soft coal, under natural basicity; (b) Smelting at 1625 °C with 10% soft coal, under natural basicity. (The yields of pig iron obtained at different smelting temperature were 20.56%, 22.97%, 24.94%, and 26.99%, respectively; The yields of pig iron obtained at different smelting duration were 23.09%, 26.96%. 30.45%, and 30.99%, respectively).

The effects of reductant dosage on the smelting separation process indices are depicted in Figure 11. The grades of the pig iron changed slightly with increasing doses (5–15%) of the reductant. However, the iron recovery rates of pig iron were obviously increased from 62.44% to 78.47%. Further increasing the reductant doses could negatively affect the values of the indices, probably due to fact Figure 10. thatFigure the 10. excessive EffectEsff ofects reductant smelting of smelting separation may separation cause parameters serious parameters on carburiz the on smelting theing smelting and separation separationlower of basicity pre of-reduced pre-reduced levels pellets of pellets. the. slag, (a) Smelting at various temperature for 20 min with 10% soft coal, under natural basicity; (b) Smelting leading(a) Smelting to the at decreased various temperature iron grade for and20 min higher with 10%viscosity soft coal, of theunder slag natural. Considering basicity; (b these) Smelting indices , the at 1625 C with 10% soft coal, under natural basicity. (The yields of pig iron obtained at different opattimal 1625 reductant °C with◦ 10% dosage soft coal, was underrecommended natural basicity as 15%.. (The yields of pig iron obtained at different smelting temperature were 20.56%, 22.97%, 24.94%, and 26.99%, respectively; The yields of pig iron smelting temperature were 20.56%, 22.97%, 24.94%, and 26.99%, respectively; The yields of pig iron obtained at different smelting duration were 23.09%, 26.96%. 30.45%, and 30.99%, respectively). obtained at different smelting duration were 23.09%, 26.96%. 30.45%, and 30.99%, respectively).

Figure 11. Eﬀects of reductant dosage on the smelting separation of pre-reduced pellets. (Smelting at

Figure1625 ◦C 11. for Effect 30 min,s of underreductant natural dosage basicity). on the (The smelting yields separation of pig iron of obtained pre-reduced at diff erentpellets reductant. (Smelting doses at 1625were °C 26.03%, for 30 26.96%, min, under 32.48%, natural and 28.17%,basicity) respectively).. (The yields of pig iron obtained at different reductant doses were 26.03%, 26.96%, 32.48%, and 28.17%, respectively). From the above result, it can be concluded that the separation of iron from the high alumina slag is not soFrom effi cientthe above due to result, the low it can recovery be concluded of iron. that Therefore, the separation some measures of iron from should the be high taken alumina to increase slag isthe not recovery so efficient rates. due to the low recovery of iron. Therefore, some measures should be taken to increaseThe the eff recoveryects of binary rates. basicity (w(CaO)/w(SiO2)) of the slag on the smelting process indices are shownThe in effect Figures of 12 binary. Increasing basicity the (w(CaO)/w(SiO basicity from natural2)) of the value slag ofon 0.0020–0.80 the smelting exerted process a littleindices impact are shownon the in iron Fig grade.ure 12. In Increasing contrast, itthe exhibited basicity afrom significant natural positive value of impact 0.0020– on0.8 the0 exerted recovery a little rates im ofpact iron with a large increase from 78.47% to 98.55%. The addition of CaO could enhance the decomposition of fayaliteFigure 11. and Effect hercynites of reductant and induce dosage the on activity the smelting of FeO, separation leading of to pre increase-reduced rates pellets of. iron (Smelting recovery, at which are1625 consistent °C for 30with min, theunder previous natural thermodynamicbasicity). (The yield calculation.s of pig iron obtained at different reductant doses were 26.03%, 26.96%, 32.48%, and 28.17%, respectively).

From the above result, it can be concluded that the separation of iron from the high alumina slag is not so efficient due to the low recovery of iron. Therefore, some measures should be taken to increase the recovery rates. The effects of binary basicity (w(CaO)/w(SiO2)) of the slag on the smelting process indices are shown in Figure 12. Increasing the basicity from natural value of 0.0020–0.80 exerted a little impact

12 of 17 on the iron grade. In contrast, it exhibited a significant positive impact on the recovery rates of iron with a large increase from 78.47% to 98.55%. The addition of CaO could enhance the decomposition of fayalite and hercynite and induce the activity of FeO, leading to increase rates of iron recovery, which are consistent with the previous thermodynamic calculation. The XRD patterns of the slag obtained from smelting separation at different basicity levels are demonstrated in Figure 13. The main mineral composition of the slag obtained from smelting separation at natural basicity is mullite. By increasing the basicity to 0.20, the content of mullite decreased, while that of corundum increased (as shown in Table 7). The peak of mullite disappeared, and the anorthite was generated when the basicity reached 0.4. Further elevating the basicity to 0.60, the XRD pattern tended to be amorphous. This is probably the reason that lots of liquid phase with lower viscosity were formed and the amorphous phase is observed after rapid cooling. The microstructures of the slags obtained from smelting separation at different binary basicity (w(CaO)/w(SiO2)) levels are presented in Figure 14 and the EDS analysis of Figure 14 is presented in Table 8. It is obvious that the main phases are metallic iron and mullite at the basicity of 0.002 (natural basicity). By increasing the basicity to 0.20, flake alumina could also can be observed, in addition to metallic iron and mullite, indicating that the separation of iron from the slag is not efficient. Further elevating the basicity to 0.40, the metallic iron disappeared and the anorthite was observed. However, there was no crystalline phase observed in the slag at basicity of 0.60 and 0.80, and the uniformity was found in the slag phase. This indicates that CaO can improve the fluidity and viscosity of slag, which are consistent with the results presented in Figure 9. All these findings agree with those of Figure 13, in which an excellent separation of iron from the slag can be achieved at 0.40–0.80 basicity. Metals 2020, 10, 57 12 of 17

Figure 12. Eﬀects of binary basicity (w(CaO)/w(SiO2)) levels on the smelting separation indices of the Figurepre-reduced 12. Effect pellets.s of binary (Reducing basicity at1625 (w(CaO)/w(SiO◦C for 30 min2)) levels and with on the 15% smelting soft coal). separation (The yields indices of pig of ironthe preobtained-reducedatdifferent pellets. (Reducing binary basicity at 1625 levels °C were for 30 32.61%, min and 33.80%, with 35.14%, 15% soft 35.18%, coal) and. (The 34.20%, yields respectively of pig iron). obtained at different binary basicity levels were 32.61%, 33.80%, 35.14%, 35.18%, and 34.20%, respectively)The XRD patterns. of the slag obtained from smelting separation at diﬀerent basicity levels are demonstrated in Figure 13. The main mineral composition of the slag obtained from smelting separation at natural basicity is mullite. By increasing the basicity to 0.20, the content of mullite decreased, while that of corundum increased (as shown in Table7). The peak of mullite disappeared, and the anorthite was generated when the basicity reached 0.4. Further elevating the basicity to 0.60, the XRD 13 of 17 pattern tended to be amorphous. This is probably the reason that lots of liquid phase with lower viscosity were formed and the amorphous phase is observed after rapid cooling.

Figure 13. Figure 13. XRDXRD patterns patterns of of the the slag slag obtained obtained from from smelting smelting separation separation at at different diﬀerent basicity basicity levels levels.. Table 7. The content of main compositions in the slag of Figure 13 (wt. %). Table 7. The content of main compositions in the slag of Figure 13 (wt. %). Basicity Iron Hercynite Mullite Corundum Anorthite Basicity Iron Hercynite Mullite Corundum Anorthite 0.0020 3 5 77 15 - 0.0020 0.203 65 577 56 3315 - - 0.20 0.406 -5 -56 - 4033 60 - 0.40 0.60------40 - - 60 0.60 0.80------0.80 - - - - -

Figure 14. Microstructures of the slags obtained from smelting separation at different basicity levels.

Table 8. The EDS analysis of Figure 14 (wt. %).

Point Fe Al2O3 SiO2 CaO 1 100 - - - 2 - 51.35 48.04 0.94 3 100 - - -- 4 - 100 - - 5 - 32.28 53.72 14

13 of 17

Metals 2020, 10, 57 13 of 17

The microstructures of the slags obtained from smelting separation at different binary basicity Figure 13. XRD patterns of the slag obtained from smelting separation at different basicity levels. (w(CaO)/w(SiO2)) levels are presented in Figure 14 and the EDS analysis of Figure 14 is presented in Table8. It is obvious that the main phases are metallic iron and mullite at the basicity of 0.002 (natural Table 7. The content of main compositions in the slag of Figure 13 (wt. %). basicity). By increasing the basicity to 0.20, flake alumina could also can be observed, in addition to metallicBasicity iron and mullite,Iron indicatingHercynite that the separationMullite of iron fromCorundum the slag is not effiAnorthitecient. Further elevating0.002 the0 basicity3 to 0.40, the metallic5 iron disappeared77 and the anorthite15 was observed.- However, there was0.20 no crystalline6 phase observed5 in the slag56 at basicity of 0.6033 and 0.80, and the uniformity- was found0.40 in the slag- phase. This indicates- that CaO- can improve the40 fluidity and viscosity60 of slag, which0.6 are0 consistent- with the results- presented in Figure- 9. All these- findings agree with- those of Figure0.8 130, in which an- excellent separation- of iron from- the slag can be- achieved at 0.40–0.80- basicity.

Figure 14. MicrostructureMicrostructuress of the slag slagss obtained from smelting separation at different diﬀerent basicity levels.levels.

Table 8. The EDS analysis of Figure 14 (wt. %). Table 8. The EDS analysis of Figure 14 (wt. %). Point Fe Al O SiO CaO Point Fe Al22O33 SiO2 2CaO 11 100100 ------22 -- 51.35 51.35 48.04 48.04 0.94 0.94 3 100 - - – 43 -100 100------54 -- 100 32.28 - 53.72- 14 65 -- 32.28 38.36 53.72 44.06 14 17.58 7 - 100 - - 8 - 48.11 38.55 13.34 9 - 36.98 39.54 23.48 10 - 36.76 38.08 25.16

3.2. Extracting Al2O3 from Smelting Separation Slag by Modifying-Alkaline Leaching The slags obtained from smelting separation at various basicity levels can be used as the burden for extracting Al2O3. The dry briquettes were fired at 900 ◦C for 120 min followed by alkaline leaching at 95 ◦C for 2 h with a liquid-to-solid ratio of 10 mL/g for the purpose of modification. Figure 15 shows the effects of the slags obtained from smelting separation at different basicity levels on the extraction yields of Al2O3. The recovery rates of Al2O3 decreased from 74.38% to 50.43% with increasing slag basicity levels from 0.0020 to 0.80. In addition, the content of Al2O3 in alkaline leaching residue increased from 8.74% to 11.35%. It can be concluded that increasing the basicity of the slag is not a suitable option for the extraction of alumina. Considering the separation of iron and extraction of alumina from the slag, the optimum basicity of the slag is recommended as 0.40. 14 of 17

6 - 38.36 44.06 17.58 7 - 100 - - 8 - 48.11 38.55 13.34 9 - 36.98 39.54 23.48 10 - 36.76 38.08 25.16

3.2. Extracting Al2O3 from Smelting Separation Slag by Modifying-Alkaline Leaching The slags obtained from smelting separation at various basicity levels can be used as the burden for extracting Al2O3. The dry briquettes were fired at 900 °C for 120 min followed by alkaline leaching at 95 °C for 2 h with a liquid-to-solid ratio of 10 mL/g for the purpose of modification. Figure 15 shows the effects of the slags obtained from smelting separation at different basicity levels on the extraction yields of Al2O3. The recovery rates of Al2O3 decreased from 74.38% to 50.43% with increasing slag basicity levels from 0.0020 to 0.80. In addition, the content of Al2O3 in alkaline leaching residue increased from 8.74% to 11.35%. It can be concluded that increasing the basicity of the slag is not a suitable option for the extraction of alumina. Considering the separation of iron and extraction of alumina from the slag, the optimum basicity of the slag is recommended as 0.40. Metals 2020, 10, 57 14 of 17

Figure 15. Eﬀects of the slags obtained from smelting separation at diﬀerent basicity levels on the

Figurerecovery 15. ratesEffect ofs of Al 2theO3. slag(Thes slagobtained was subjectedfrom smelting to firing separation at 900 ◦ Cat for different 120 min basicity and leaching levels aton 95 the◦C recoveryfor 120 min,rates with of Al a2O liquid-to-solid3. (The slag was ratio subjected of 10 mL to/g firing at 300 at rpm). 900 °C (The for Al1202O min3 contents and leaching of modified at 95 slag°C forobtained 120 min at, with different a liquid basicity-to-solid levels ratio were of 10 17.11%, mL/g at 17.88%, 300 rpm 18.51%,). (The 19.27%, Al2O3 content 19.72%,s of respectively. modified slag The obtainedyields of at leaching different residues basicity obtained levels w atere di 17.11%,fferent basicity17.88%, levels18.51%, were 19.27%, 50.16%, 19.72%, 54.16%, respectively. 58.86%, 71.32%, The yieldands 86.13%, of leaching respectively). residues obtained at different basicity levels were 50.16%, 54.16%, 58.86%, 71.32%, and 86.13%, respectively.). 3.3. Fe and Al Balance in the Full Flow Sheet 3.3. Fe and Al Balance in the Full Flow Sheet As a whole, the following conditions were optimized: pre-reduction at 1000 ◦C for 60 min with a C/AsFe a mass whole ratio, the of following 1.5, smelting conditions the pre-reduced were optimized pellets: topre recover-reduction iron at at 1000 1625 °C◦C for for 60 30 min min with with a15% C/Fe soft mass coal ratio at 0.4 of binary1.5, smelting basicity the (w(CaO) pre-reduced/w(SiO 2pellets)), then to roasting recover theiron slag at 1625 at 900 °C◦C for for 30 120 min min with and 15%alkaline soft coal leaching at 0.4 thebinary slag basicity to extract (w(CaO)/w(SiO alumina at 952)),◦ Cthen for roasting 120 min the with slag a liquid-to-solidat 900 °C for 120 of min 10 mLand/g. alkalineA full flow leaching sheet the of theslag exergy to extract balance alumina of Fe at and 95 Al°C elementsfor 120 min is illustratedwith a liquid in- Figureto-solid 16 of. 10 mL/g. A full flowTable sheet9 summarizes of the exergy the balance chemical of Fe compositions and Al elements of pig is iron illustrated and alkaline in Figure leaching 16. residue. It can be seenTable that 9 summarizes the pig iron the yielded chemical 97.08% compositions Fetotal with of pig an iron overall and iron alkaline recovery leaching of 94.54%, residue. and It can the bextractione seen that rate the of pig Al 2 ironO3 by yielded alkaline 97.08% leaching Fetotal was with 68.93%. an overall The pig iron iron recovery can be applied of 94.54% as the, and burden the extractionfor steelmaking rate of byAl2 anO3 electricby alkaline arc furnace,leaching whilewas 68.93%. the pure The Al 2pigO3 ironcan becan manufactural be applied as from the leachingburden forsolution steelmaking and the by alkaline an electric leaching arc furnace, residue canwhile be the used pure to produceAl2O3 can cement be manufactural [26,27]. Effective from utilizationleaching of HAIO could be achieved through the proposed method.

Table 9. Chemical compositions of pig iron and alkaline leaching tailing (wt. %).

Sample Fe Al2O3 SiO2 CaO MgO K2O Na2OPS Pig iron 97.08 0.13 0.10 0.025 0.0074 0.0029 0.0045 0.28 0.14 Alkaline leaching residue 3.4 9.77 20.93 31.53 0.38 0.11 3.73 0.84 0.015

Baev et al. [28] investigated the heat exchange and reduction in high-alumina charges in the Blast furnace. The results showed that, due to the lower reducibility of HAIO, the coke and energy consumption were up to 1200 kg/(ton of pig iron) and 460 kWh/(ton of pig iron) in the blast furnace smelting process, respectively. He et al. [29] also reported that the coke and energy consumption of the direct slag-iron smelting reduction were 1463.30 kg/(ton of pig iron) and 432.29 kWh/(ton of pig iron), respectively. The energy and reductant consumption of the pre-reduction-smelting separation of HAIO treatment were decreased with increasing metallization rates of pre-reduced pellets [30]. According to a previous report [30], it can be calculated that the energy and reductant consumption of this process is 380 kWh/(ton of pig iron) and 1014 kg/(ton of pig iron), respectively. 15 of 17

Metalssolution2020 and, 10, 57the alkaline leaching residue can be used to produce cement [26,27]. Effective utilization15 of 17 of HAIO could be achieved through the proposed method.

Figure 16. A complete flow ﬂow sheet for the exergy balanc balancee of Fe and Al elements.elements.

4. ConclusionsTable 9. Chemical compositions of pig iron and alkaline leaching tailing (wt. %). In this study, an innovative technique, consisting of pelleting-pre-reducing-smelting Sample Fe Al2O3 SiO2 CaO MgO K2O Na2O P S reduction-modifying-alkaline leaching steps, was developed for extraction of Fe and Al2O3 from HAIO. ConclusionsPig can iron be drawn as follows:97.08 0.13 0.10 0.025 0.0074 0.0029 0.0045 0.28 0.14 Alkaline leaching residue 3.4 9.77 20.93 31.53 0.38 0.11 3.73 0.84 0.015 1. HAIO, assaying 33.43% Fetotal, 19.09% Al2O3, and 18.36% SiO2, was defined as a refractory iron Baevore and et al. employed [28] investigated as the raw the materials heat exchange to yield and pig reduction iron and extractin high Al-alumina2O3. In charges addition, in the the major Blast furnacirone. The minerals results of HAIO showed are that hematite, due and to the goethite, lower while reducibility the main of aluminum HAIO, the minerals coke are and kaolinite, energy consumptionand all of were them up are toclosely 1200 kg/ associated(ton of pig at iron superfine) and 460 size. kWh/(ton of pig iron) in the blast furnace 2.smeltingApproximately process, respectively. 94.54% of FeHe was et al. recovered [29] also inreported the pig that iron the when coke smelting and energy the pre-reduced consumption high of the directaluminum slag-iron iron smelting pellets atreduction 1625 ◦C were for 30 1463.30 min with kg/(ton a slag of basicitypig iron) of and 0.40, 432.29 while kWh approximately/(ton of pig iron),68.93% respectively Al2O3. Thewas energy extracted and from reductant the slag consumption by roasting theof the smelting pre-reduction separation-smelting slag at separation 900 ◦C for of HAIO2 hand treatment alkaline were leaching decreased at 95 ◦ withC for increasing 2 h with a Lmetallization/S ratio of 10 rate mL/sg, of and pre the-reduced leaching pellets residue [30]. is Accordingsuitable to fora previous cement industry.report [30], it can be calculated that the energy and reductant consumption 3.of thisThe process developed is 380 processkWh/(ton can of provide pig iron) an and alternative 1014 kg/(ton for eff ofective pig iron) and, green respectively. utilization of HAIO.

4. Conclusions Author Contributions: Funding acquisition, J.P. and Z.G.; investigation, S.L., Y.S., J.C., and J.X.; methodology, J.P. andIn D.Z.;this study, supervision, an innovative J.P., D.Z., technique, and Z.G.; writing—original consisting of pelleting draft, S.L.;-pre writing—review-reducing-smelting and editing,reduction S.L.- All authors have read and agreed to the published version of the manuscript. modifying-alkaline leaching steps, was developed for extraction of Fe and Al2O3 from HAIO. Funding:ConclusionsThis can study be was drawn funded as byfollows: the Nation Natural Science Foundation of China (grant number NO. 51574281), the Youth natural science foundation China (grant number NO. 51904347) and Innovation-driven Project of total 2 3 2 Guangxi1. HAIO Zhuang, assaying Autonomous 33.43% RegionFe , (grant19.09% number Al O , No. and AA18242003, 18.36% SiO No., was AA148242003). defined as a refractory iron ore and employed as the raw materials to yield pig iron and extract Al2O3. In addition, the major Conﬂicts of Interest: The authors declare no conﬂict of interest. iron minerals of HAIO are hematite and goethite, while the main aluminum minerals are kaolinite, and all of them are closely associated at superfine size

Metals 2020, 10, 57 16 of 17

References

1. Zhang, K.Y.; Kleit, A.N. Mining rate optimization considering the stockpiling: A theoretical economics and real option model. Resour. Policy 2016, 47, 87–94. [CrossRef] 2. Pradhan, N.; Das, B.; Gahan, C.S.; Kar, R.N.; Sukla, L.B. Beneficiation of iron ore slime using aspergillus niger and bacillus circulans. Bioresour. Technol. 2006, 97, 1876–1879. [CrossRef][PubMed] 3. Singh, S.; Sahoo, H.; Rath, S.S.; Sahu, A.K.; Das, B. Recovery of iron minerals from Indian iron ore slimes using colloidal magnetic coating. Powder Technol. 2015, 269, 38–45. [CrossRef] 4. Lu, L.; Holmes, R.J.; Manuel, J.R. Effects of alumina on sintering performance of hematite iron ores. ISIJ Int. 2007, 47, 349–358. [CrossRef] 5. Umadevi, T.; Prakash, S.; Bandopadhyay, U.K.; Mahapatra, P.C.; Prabhu, M.; Ranjan, M. Influence of alumina on iron ore sinter quality and productivity. World Iron Steel 2010, 10, 12–18. 6. Rocha, L.; Cançado, R.L.; Peres, A.C. Iron ore slimes flotation. Miner. Eng. 2010, 23, 842–845. [CrossRef] 7. Thella, J.S.; Mukherjee, A.K.; Srikakulapu, N.G. Processing of high alumina iron ore slimes using classification and flotation. Powder Technol. 2012, 217, 418–426. [CrossRef] 8. Shobhana, D.; Santosh, P.; Ratnakar, S.; Gayna, M.P. Response of process parameters for processing of iron ore slime using column flotation. Int. J. Miner. Process 2015, 140, 58–65. 9. Mahiuddin, S.; Bondyopadhway, S.; Baruah, J.N. A study on the beneficiation of indian iron-ore fines and slime using chemical additives. Int. J. Miner. Process 1989, 26, 285–296. [CrossRef] 10. Shu, C.W.; Tang, Y.; Wang, Z.Q.; Zhang, Q. Research on sodium salt roasting-acid leaching technology for high-alumina and high-silicon refractory limonite. Min. Metall. Eng. 2012, 32, 62–65. 11. Zhu, D.Q.; Chun, T.J.; Pan, J. Recovery of Iron from High-Iron Red Mud by Reduction Roasting with Adding Sodium Salt. J. Iron. Steel Res. Int. 2012, 19, 1–5. [CrossRef] 12. Chun, T.J.; Zhu, D.Q.; Pan, J. Preparation of metallic iron powder from red mud by sodium salt roasting and magnetic separation. Can. Metall. Q. 2014, 53, 183–189. [CrossRef] 13. Zhou, X.L.; Zhu, D.Q.; Pan, J.; Luo, Y.H.; Liu, X.Q. Upgrading of High-Aluminum Hematite-Limonite Ore by High Temperature Reduction-Wet Magnetic Separation Process. Metals 2016, 6, 57. [CrossRef] 14. Liu, Y.; Zuo, K.; Yang, G.; Shang, Z.; Zhang, J. Recovery of ferric oxide from bayer red mud by reduction roasting-magnetic separation process. J. Wuhan Univ. Technol. Mater. Sci. Ed. 2016, 31, 404–407. [CrossRef] 15. Chun, T.J.; Long, H.M.; Li, J.X. Alumina-Iron Separation of High Alumina Iron Ore by Carbothermic Reduction and Magnetic Separation. Sep. Sci. Technol. 2015, 50, 760–766. [CrossRef] 16. Sellaeg, H.; Kolbeinsen, L.; Safarian, J. Iron separation from bauxite through smelting-reduction process. Light Metals. 2017, 127–135. 17. He, Y.B.; Tang, B.; Li, Q.; Zou, Z.S. A comprehensive static model of an iron bath smelting reduction process for alumina-rich iron ore. ISIJ Int. 2015, 55, 2135–2144. [CrossRef] 18. Guo, Z.Q.; Zhu, D.Q.; Pan, J.; Zhang, F. Innovative methodology for comprehensive and harmless utilization of waste copper slag via selective reduction-magnetic separation process. J. Clean. Prod. 2018, 187, 910–922. [CrossRef] 19. Yang, C.C.; Zhu, D.Q.; Pan, J.; Lu, L.M. Simultaneous Recovery of Iron and Phosphorus from a High-Phosphorus Oolitic Iron Ore to Prepare Fe-P Alloy for High-Phosphorus Steel Production. JOM 2017, 69, 1663–1668. [CrossRef] 20. Li, S.W.; Pan, J.; Zhu, D.Q.; Guo, Z.Q.; Xu, J.W.; Chou, J.L. A novel process to upgrade the copper slag by direct

reduction-magnetic separation with the addition of Na2CO3 and CaO. Powder Technol. 2019, 347, 159–169. [CrossRef] 21. Qin, S.C.; Jiang, K.X.; Zhang, B.S.; Wang, H.B. Study on phase form and content of sulfur and iron in oxygen pressure acidic leaching residue of sphalerites. Chin. J. Inorg. Anal. Chem. 2016, 6, 57–61. 22. Wang, W.H.; Wang, D.; Li, X.Y. Phase analysis on aluminum metal and aluminum trioxide in aluminium ash. Resour. Environ. Eng. 2009, 23, 33–334. 23. Iron Ores—Determination of Metallic Iron Content—Sulfosalicylic Acid Spectrophotometric Method; GB/T 24194-2009; Wuhan lron and Steel Company: Wuhan, China, 2010. 24. Yang, Y.X.; Qi, X. X-ray Diﬀraction Analysis; Shanghai Jiaotong University Press: Shanghai, China, 1989; pp. 270–278. Metals 2020, 10, 57 17 of 17

25. Zhang, F.; Zhu, D.Q.; Pan, J.; Guo, Z.Q. Eﬀect of basicity on the structure characteristics of chromium-nickel bearing iron ore pellets. Powder Technol. 2019, 342, 409–417. [CrossRef] 26. Gu, X.; Dong, W.; Chen, Y.; Li, Y. Investigation on the glass-anorthite insulating composites with low dielectric constant and low sintering temperature. Rare Metal Mater. Eng. 2007, 36, 464–467. 27. Liu, Q.; Pan, Z.; Li, Q.; Ruan, Y. Preparation of anorthite lightweight thermal insulating brick and the formation process of anorthite. Bull. Chin. Ceram. Soc. 2010, 29, 1269–1274. 28. Baev, S.V.; Uryavin, G.A.; Leont’ev, L.N. Heat exchange and reduction in high-alumina charges in the blast furnace. Metallurgist 1975, 19, 351–354. [CrossRef] 29. He, Y.B. Study on Slag-Iron Bath Smelting Reduction Process for Alumina-Rich Iron Ore Processing; Northeastern University: Shenyang, China, 2016; pp. 24–34. 30. Gao, J.J.; Wang, X.Y.; Qi, Y.H.; Wang, F. Technical analysis on ironmaking process of rotary kiln pre-reduction and smelting by coal and oxygen. J. Iron Steel Res. 2018, 30, 91–96.

© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).