Preparation of a Master Fe–Cu Alloy by Smelting of a Cu-Bearing Direct Reduction Iron Powder

metals

Article Preparation of a Master Fe–Cu Alloy by Smelting of a Cu-Bearing Direct Reduction Iron Powder

Liaoting Pan, Deqing Zhu, Zhengqi Guo * and Jian Pan School of Mineral Processing and Bioengineering, Central South University, Changsha 410083, China; [email protected] (L.P.); [email protected] (D.Z.); [email protected] (J.P.) * Correspondence: [email protected]; Tel.: +86-185-7313-1417; Fax: +86-731-8883-6942

Received: 13 May 2019; Accepted: 18 June 2019; Published: 21 June 2019

Abstract: Generally, the Cu-bearing direct reduction iron powder (CBDRI) obtained from a direct reduction-magnetic separation process of waste copper slag contains a high content of impurities and cannot be directly used to produce Cu-bearing special steel. In this paper, further smelting treatment of CBDRI was conducted to remove its impurities (such as S, SiO2, Al2O3, CaO and MgO) and acquire a high-quality Fe–Cu master alloy. The results show that the Fe–Cu master alloy, assaying 95.9% Fe, 1.4% Cu and minor impurities, can be obtained from the smelting process at 1550 ◦C for 40 min with 1.0 basicity. Meanwhile, the corresponding iron and copper recoveries are 98.6% and 97.2%, respectively. Theoretical calculations and experimental results show that appropriate basicity (0.9~1.1) is beneficial for the recovery of Fe and Cu from a thermodynamic viewpoint due to the excellent fluidity of the slag in this basicity range. Moreover, the mechanism of desulfurization was revealed by calculating the sulfide capacity and the desulfurization reaction kinetics. Increasing the binary basicity of the slag benefits both the sulfide capacity and diffusion coefficient of the sulfur in the molten slag, resulting in higher desulfurization efficiency and lower S content in the master alloy.

Keywords: Cu-bearing direct reduction iron; smelting; Cu-bearing steel; basicity; sulﬁde capacity; desulfurization

1. Introduction Cu-bearing steel is characterized by high strength, excellent corrosion resistance and high antibacterial ability [1–4]. Based on this characterization, many kinds of Cu-bearing steels, such as weathering steel, antibacterial stainless steel and high-strength steel (HSAL serials), have been rapidly developed in recent years [5–7]. Generally, conventional Cu-bearing steel production methods require the addition of electrolytic copper to adjust the copper content in the products [8–10]. Unfortunately, with the depletion of high-grade copper mineral resources in the world, the production costs for electrolytic copper remains high. This problem could ultimately lead to higher production costs for Cu-bearing steel. However, low-cost ferroalloy (Fe–Cu alloy) may replace electrolytic copper in the production of the Cu-bearing steels, which provides the possibility of innovation lowering technology costs. Copper smelting slag typically contains 40% Fe and 1% Cu and can be considered as an important secondary resource for utilization [11]. Recently, researchers have proposed processes to recover valuable metals from copper slag [12–17]. However, those processes mainly focused on Cu recovery while totally neglecting Fe recovery. Direct reduction process, as a novel process, was developed to treat the copper slag [8,10,18–22]. In this process, the iron oxides and copper bearing minerals were reduced to metallic iron and copper. The metallic copper dissolves into metallic Fe lattices to form a Fe–Cu solid solution (Fe–Cu alloy). The Fe–Cu solid solution can then undergo subsequent grinding and magnetic separation processes. Therefore, this process can fulﬁl the comprehensive recovery of Fe and Cu from copper slag and attain Fe–Cu direct reduction iron powder (CBDRI). CBDRI could

Metals 2019, 9, 701; doi:10.3390/met9060701 www.mdpi.com/journal/metals Metals 2 of 18 subsequent grinding and magnetic separation processes. Therefore, this process can fulfil the comprehensiveMetals 2019, 9, 701 recovery of Fe and Cu from copper slag and attain Fe–Cu direct reduction 2 iron of 13 powder (CBDRI). CBDRI could potentially substitute for scrap steel and electrolytic copper in the production of Cu-bearing steel and may lead to lower costs for steel production. However, most CBDRIpotentially is not substitute suitable for for scrap being steel directly and electrolytic used in copper steel production in the production due to of its Cu-bearing high impurity steel and may lead to lower costs for steel production. However, most CBDRI is not suitable for being concentration (SiO2, Al2O3, CaO, MgO and S), therefore, a necessary extra smelting step to remove thosedirectly impurities used in steel and production obtain a due higher to its highFe–Cu impurity alloy concentrationpurity is indispensable. (SiO2, Al2O 3, There CaO, MgO are few and investigations,S), therefore, a however, necessary on extra CBDRI smelting for the step production to remove of those Cu- impuritiesbearing steel and master obtain alloy a higher via electric Fe–Cu furnacealloy purity smelting. is indispensable. There are few investigations, however, on CBDRI for the production of Cu-bearingCBDRI steelhas been master successfully alloy via electricprepared furnace from copper smelting. smelting slag by a direct reduction-magnetic separationCBDRI process has been as reported successfully in our prepared earlier frompaper copper [10]. The smelting aim of slag thisby research a direct is reduction-magnetic to produce a high- qualityseparation Fe–Cu process master as alloy reported from in CBDRI our earlier by a smelting paper [10 proces]. Thes aimand ofdirectly this research use the ismaster to produce alloy as a thehigh-quality raw material Fe–Cu to master produce alloy Cu from-bearing CBDRI steel by a bysmelting an electric process arc and furnace directly to use replace the master part alloy of the as electrolyticthe raw material copper to and produce scrap Cu-bearing steel. Additionally, steel by an the electric desulfurization arc furnace tobehavior replace and part its of mechanism the electrolytic in thecopper smelting and scrap process steel. were Additionally, revealed in the this desulfurization work. behavior and its mechanism in the smelting process were revealed in this work. 2. Experimental 2. Experimental 2.1. Raw Materials 2.1. Raw Materials The Fe–Cu direct reduction iron powder (CBDRI)used in this study was obtained from the direct The Fe–Cu direct reduction iron powder (CBDRI)used in this study was obtained from the direct reduction and magnetic separation of copper smelting slag. The chemical mass composition of CBDRI reduction and magnetic separation of copper smelting slag. The chemical mass composition of CBDRI shown in Table 1 and demonstrate that CBDRI contains 90.33% Fe and 1.29% Cu, which are potential shown in Table1 and demonstrate that CBDRI contains 90.33% Fe and 1.29% Cu, which are potential materials for the preparation of copper bearing steel. Some undesirable impurities, such as SiO2, materials for the preparation of copper bearing steel. Some undesirable impurities, such as SiO2, Al2O3, CaO, and MgO, are also found in the CBDRI. In particular, the S content in CBDRI is as high Al2O3, CaO, and MgO, are also found in the CBDRI. In particular, the S content in CBDRI is as high as as 0.3%, which is extremely harmful to the production of Cu-bearing steel. The smelting treatment of 0.3%, which is extremely harmful to the production of Cu-bearing steel. The smelting treatment of the the CBDRI is necessary to ensure the quality of the alloy. Furthermore, the size distribution of CBDRI CBDRI is necessary to ensure the quality of the alloy. Furthermore, the size distribution of CBDRI presented in Figure 1 shows that the CBDRI is very fine, with over 90% below 0.074 mm. presented in Figure1 shows that the CBDRI is very ﬁne, with over 90% below 0.074 mm.

Table 1. Chemical compositions of CBDRI (wt %). Table 1. Chemical compositions of CBDRI (wt %).

Fe FeCu Cu S SMFe MFe SiO SiO2 2 AlAl22OO33 CaOCaO MgO NaNa2OCP2O C P 90.3390.33 1.29 1.29 0.30 0.30 87.54 87.54 3.62 3.62 0.56 0.56 1.09 1.09 0.29 0.070.07 0.780.78 0.0120.012

10 Particle Particle size distribution/%

-0.037 0.037-0.043 0.043-0.074 0.074-0.15 +0.15 Size/mm Figure 1. SizeSize distribution distribution of of CBDRI CBDRI.. The chemical composition of limestone is shown in Table2. Limestone containing 55.73% of CaO was used as both a ﬂux and desulfurizer to adjust the binary basicity (ratio of CaO/SiO ) and remove 2 sulfur in the smelting process. Its particle size was less than 0.074 mm. Metals 3 of 18

The chemical composition of limestone is shown in Table 2. Limestone containing 55.73% of CaO Metalswas used2019, as9, 701both a flux and desulfurizer to adjust the binary basicity (ratio of CaO/SiO2) and remove3 of 13 sulfur in the smelting process. Its particle size was less than 0.074 mm.

Table 2. Chemical compositions of limestone(wt %). Table 2. Chemical compositions of limestone(wt %).

CaO Al2O3 SiO2 MgO P S MnO LOI CaO Al2O3 SiO2 MgO P S MnO LOI 55.73 55.731.19 1.194.13 4.132.05 2.05 0.0010.001 0.0010.001 0.020.02 41.52 41.52

2.2. Experimental Methods

2.2.1. Smelting Process

The smelting tests were performed under a controlled environment in aa verticalvertical MoSiMoSi22 heating furnace (schematic diagram of the smelting furnace is shown in FigureFigure2 2).). TheThe proportionproportion integralintegral didifferentialﬀerential (PID)temperature(PID)temperature control programme in the furnace was used to limit the temperature ﬂuctuation range to 5 C. In each test, 30 g of CBDRI were mixed uniformly with the required fluctuation range to ± ±5 °C.◦ In each test, 30 g of CBDRI were mixed uniformly with the required amount of limestone. The The mixtures mixtures were then loaded into a 50 mLmL corundumcorundum crucible.crucible. When the furnace temperature was raised to the smelting temperature (1515~1600(1515~1600 °C◦C),), the crucible was quickly placed into the heatingheating zone of the furnace and smelted for a given period of time (10~60min)(10~60min) under a N N22 atmosphere.atmosphere. At At the the end end of smelting, of smelting, the crucible the crucible was taken was taken out of out the offurnace the furnace and quickly and quickly cooled cooledunder the under cover the via cover coke via breeze. coke breeze.Finally, Finally,the crucible the crucible was broken was brokento separa tote separate and weigh and the weigh Fe– theCu Fe–Cualloy and alloy slag and carefully slag carefully for sample for sample preparation. preparation. The The samples samples were were then then prepared prepared to to assay assay the chemical compositions.

Thermocouple

Refractory

MoSi2 Heating element

Crucible

Ar gas

Figure 2. Schematic diagram of the smelting furnace furnace..

The recovery rate of iron and Cu η was calculated fromfrom EquationEquation (1):(1):

M푀1 × TM푇푀1 η휂== 1 × 1 × 100%100% (1 (1)) M푀0 × TM푇푀0 × 0 × 0 Where η is the recovery rate of Fe or Cu; M1 is the mass of the Fe–Cu master alloy; TM1 is the where η is the recovery rate of Fe or Cu; M1 is the mass of the Fe–Cu master alloy; TM1 is the grade of grade of Fe or Cu in the master alloy; M0 is the mass of CBDRI; TM0 is the grade of Fe or Cu in CBDRI. Fe or Cu in the master alloy; M0 is the mass of CBDRI; TM0 is the grade of Fe or Cu in CBDRI.

2.2.2. Analytic Tests A thermodynamic package FactSage (Version (Version 7.0) software ( (GTT-Technologies,GTT-Technologies, Herzogenrath Herzogenrath,, Germany) was used to calculate the phase diagram of Fe–Cu-C, SiO2-Al2O3-CaO-MgO and the Germany) was used to calculate the phase diagram of Fe–Cu-C, SiO2-Al2O3-CaO-MgO and the viscosity of the molten slag. The phase phase compositions of the samples samples were investig investigatedated using an X X-ray-ray di ﬀractometer (XRD, D/max 2550 PC, Rigaku Co., Ltd, Tokyo, Japan). Microstructure analyses of Metals 4 of 18 diffractometerMetals 2019, 9, 701 (XRD, D/max 2550 PC, Rigaku Co., Ltd, Tokyo, Japan). Microstructure analyses4 of of 13 master alloy were observed by a Leica DMLP optical microscopy, FEI Quata-200 scanning electron microscopemaster alloy (SEM, were TESCAN, observed byMIRA3 a Leica LMU, DMLP Brno optical, Czech microscopy, Republic) and FEI Quata-200an EDAX energy scanning dispersive electron Xmicroscope-ray spectrometer (SEM, TESCAN,(EDS). The MIRA3 SEM images LMU, Brno,were recorded Czech Republic) in backscatter and an electron EDAX energymodes dispersiveoperating inX-ray the low spectrometer vacuum mode (EDS). at The 0.5 SEMTorr imagesand 20 keV. were recorded in backscatter electron modes operating in the low vacuum mode at 0.5 Torr and 20 keV. 3. Results and Discussion 3. Results and Discussion 3.1. Thermodynamics Calculations of Smelting Process 3.1. Thermodynamics Calculations of Smelting Process The thermodynamics calculations for the molten slag and alloy in smelting process were conductedThe thermodynamics by FactSage 7.0 calculationsto analyse the for feasibility the molten for slag separation and alloy between in smelting the process slag and were master conducted alloy. Theby FactSage quaternary 7.0 to phase analyse diagram the feasibility of the forCaO separation–MgO–SiO between2–Al2O3 the primary slag and slag master is presented alloy. The in quaternary Figure 3 andphase shows diagram that ofthe the primary CaO–MgO–SiO slag possesses2–Al2O 3theprimary appropriate slag is liquidus presented temperatu in Figure3re and at approximately shows that the 1450primary °C. slag possesses the appropriate liquidus temperature at approximately 1450 ◦C.

SiO2

T oC Slag-liq#2 1800

SiO2

1 1700 SiO2 5

Chemical 7

18 1600 component Mg2Al4Si5O18 1920 24 area of slag Clinopyroxene 9 Wollastonite, 1500

3 22 Mullite 23 21 CaAl2Si2 O8 1400

8 13 3 Ca3Si2 O117 10 1300 Melilite 1712 4

24 1200 CaMg2Al16 O27 (s)

a-Ca SiO 2 4 6 1100 Al2O3 2

1 Monoxide 1000 Spinel Monoxide

1516 14 CaO 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 Al O mass fraction 2 3

FigureFigure 3.3. Phase diagram of SiO22--AlAl2OO33--CaOCaO with with MgO MgO == 5.22%5.22%..

InIn addition, addition, the the liquid liquid projection projection of of the the ternary ternary alloy alloy was was carried carried out; out; the the results results are are shown shown in in FigureFigure 44.. TheThe liquidusliquidus temperaturetemperature of of this this ternary ternary alloy alloy was was near near 1450 1450◦ C.°C. Based Based on on the the Figures Figs. 33 andand 44,, thethe smelting smelting temperature temperature should be 1500~1550 °C◦C to to ensure ensure a a good good separation separation between between alloy alloy and and sla slagg andand to to keep keep a a superheated superheated temperature range. 3.2. Smelting Process of CBDRI 3.2. Smelting Process of CBDRI As shown in Table1, iron and copper, as the main valuable metals in DRI, should be considered As shown in Table 1, iron and copper, as the main valuable metals in DRI, should be considered ﬁrst for recovery. Therefore, the experiments on the eﬀect of smelting conditions on Fe and Cu recovery first for recovery. Therefore, the experiments on the effect of smelting conditions on Fe and Cu and quality of the master alloy were carried out systematically. recovery and quality of the master alloy were carried out systematically.

Metals 2019, 9, 701 5 of 13 Metals 5 of 18 Metals 5 of 18

2.0 2.0 Fe - Cu - C Fe - Cu - C Projection (LIQUID), 1 atm 1.8 1400 oC Projection (LIQUID), 1 atm 1.8 1400 oC 1.6 1.6

1.4

)/% )/%

C 1.2 +

C 1.2 1450 oC + 1450 oC

Cu 1.0 +

Fe Chemical component Fe /( Chemical component

/( 0.8 C 0.8 of alloy C of alloy o 0.6 1500 oC 0.6 1500 C o 1550 oC o 1 1550 C 0.4 1500 oC 1 1500 C o 0.4 1600 C 1600 oC 0.2 0.2 0 0 0 0.5 1.0 1.5 2.0 2.5 3.0 0 0.5 1.0 1.5 2.0 2.5 3.0 Cu/(Fe+Cu+C)/% Cu/(Fe+Cu+C)/% Figure 4. Phase diagram of Fe–Cu-C. Figure 4. Phase diagram of Fe–Cu-C.Fe–Cu-C.

3.2.1. Effect Effect of Smelting Temperature Undoubtedly, smeltingsmelting temperaturetemperature has has a a significant significant e ffeffectect on on the thethe smelting smeltingsmelting process. process.process. The TheThe eff effectecteffect of ofthe the smelting smelting temperature temperature on Fe on and Fe Cu and recovery Cu recovery and the quality and the of quality the master of thealloy master was investigated. alloy was Theinvestigated.investigated. results (Figure TheThe results5results) show ((Figure that with 5) anshow increase that with in the an smelting increase temperature, in the smelting the Fe temperature, and Cu recovery the Fe is andelevated Cu recovery and maintained is elevated approximately and maintained 96%. Additionally,approximately the 96%. S content Additionally, of the master the S alloy content decreased of the mastersharply alloy and wasdecreased maintained sharply at 0.13%.and was When maintained the smelting at 0.13%. temperature When the was smelting over 1550 temperature◦C, all indexes was overslightly 1550 changed. °C, all indexes slightly changed.

100 0.30 100 0.30 98 98 98 Fe 98 97 Fe 0.25 97 0.25 96 96 96 96 TFe 0.20 Cu TFe 94 0.20 Cu 94 95 95 92 92 0.15 94 TS 0.15 94 TS 90 1.890 content/% S 93 1.8 content/% S 93 1.6 0.10

Fe Fe Cu and grade/% 1.6 0.10

Fe Fe Cu and grade/% 1.4 Fe Fe Cu and recovery/% 1.4

Fe Fe Cu and recovery/% 92 92 TCu 1.2 TCu 1.2 1.0 0.05 1.0 0.05 91 Fe recovery Fe grade of alloy 91 Fe recovery Fe grade of alloy S content of alloy 0.8 Cu recovery Cu grade of alloy S content of alloy 0.8 Cu recovery Cu grade of alloy 0.6 90 0.6 0.00 90 0.00 1525 1550 1575 1600 1525 1550 1575 1600 Smelting temperature /C Smelting temperature /C Figure 5. EffectEffect of smeltingsmelting temperature on Fe and CuCu recovery and S removal (Smelting for 30 min Figure 5. Effect of smelting temperature on Fe and Cu recovery and S removal (Smelting for 30 min with 0.7 binary basicity).basicity). with 0.7 binary basicity). Generally, high temperatures are beneficial for improving the fluidity and decreasing the Generally, high temperatures are beneficial for improving the fluidity and decreasing the viscosity of molten slag, thus, resulting in a better settling and separation of the alloy particles. viscosity of molten slag, thus, resulting in a better settling and separation of the alloy particles. Therefore, the recovery of Fe and Cu increased as the temperature increased. In addition, Therefore, the recovery of Fe and Cu increased as the temperature increased.. In In addition, addition, the the the desulfurization reaction in the smelting process is an endothermic reaction, and higher temperatures desulfurization reaction in the smelting process is an endothermic reaction, and higher temperatures are favourable for the desulfurization in view of the thermodynamics principles. Furthermore, based are favourable for the desulfurization in view of the thermodynamics principles.. Furthermore,Furthermore, basedbased on dynamics theory, higher temperatures are also capable of promoting the diffusion of S2 2−and O22− on dynamics theory, higher temperaturesres areare alsoalso capablecapable ofof promotingpromoting thethe diffusiondiffusion ofof SS−2− and O−2− ions from an alloy phase to a slag phase, and ultimately improve the desulfurization efficiency [23]. ionsions fromfrom anan alloyalloy phasephase toto aa slagslag phase,phase, andand ultimatelyultimately improveimprove thethe desulfurizationdesulfurization efficiency [2[23]].. Therefore, the suitable smelting temperaturetemperature forfor thethe smeltingsmelting ofof CBDRICBDRI isis suggestedsuggested atat 15501550 °C, which

Metals 2019, 9, 701 6 of 13

Metals 6 of 18 Therefore, the suitable smelting temperature for the smelting of CBDRI is suggested at 1550 ◦C, which is in accordance with the thermodynamicsthermodynamics analysis.

3.2.2. Effect Effect of Smelting Duration Smelting duration isis believedbelieved toto havehave a a significant significant e ffeffectect on on the the Fe Fe and and Cu Cu recovery recovery and and the the Fe–Cu Fe– Cumaster master alloy alloy quality. quality. Figure Figure6 shows 6 shows the etheffect effect of the of smeltingthe smelting duration duration on the on Fethe and Fe and Cu gradeCu grade and andrecovery recovery and and the removalthe removal of S. of As S. the As smeltingthe smelting duration duration increased increased from from 10 min 10 min to 40 to min, 40 min, the Fethe and Fe andCu recovery Cu recovery increased increased from from 94.4% 94.4% and 93.8%and 93.8% to 98.2% to 98.2% and 95.8%,and 95.8%, respectively. respectively. Correspondingly, Correspondingly, the S thecontent S content of the Fe–Cu of the master Fe–Cu alloy master presented alloy a decreasing presented trend a decreasing from 0.18% trend to 0.1%. from Furthermore, 0.18% to the 0.1%. Fe Furthermore,and Cu grade the of the Fe masterand Cu alloy grade changed of the master slightly. alloy When changed the smelting slightly. duration When wasthe smelting further prolonged, duration wasthe Fefurther and Cu prolonged, recovery the and Fe S contentand Cu wererecovery improved and S content insignificantly. were improved insignificantly.

100 0.30 98 98 97 Fe 0.25 TFe 96 96 0.20 Cu 94 95 92 0.15 94 TS 90

1.8 S content/% 93 1.6 0.10 Fe and Cu grade/% 1.4 Fe and Cu recovery/% 92 TCu 1.2 1.0 0.05 91 Fe recovery Fe grade S content 0.8 Cu recovery Cu grade 0.6 90 0.00 0 10 20 30 40 50 60 70 Smelting duration/min

Figure 6. EffectEﬀect of of smelting smelting time time on on Fe Fe and and Cu Cu recovery recovery and and S removal S removal (Smelti (Smeltingng at 1550 at 1550 °C with◦C with 0.7 basicity)0.7 basicity)..

These results imply that the insuinsufficientﬃcient smelting time cannot ensure that the desulfurization reaction andand thethe settling settling of of the the Fe–Cu Fe–Cu master master alloy alloy will will carry carry out out thoroughly, thoroughly, thus, thus, leading leading to low to metal low metalrecovery recovery and poor and alloypoor quality.alloy quality. Therefore, Therefore, 40 min 40 is min appropriate is appropriate for the for simultaneous the simultaneous recovery recovery of Fe ofand Fe Cu and and Cu the and removal the removal of S. of S.

3.2.3. Effect of Binary Basicity 3.2.3. Effect of Binary Basicity The influence of binary basicity on Fe and Cu recovery and the removal of S illustrated in Figure7 The influence of binary basicity on Fe and Cu recovery and the removal of S illustrated in Figure and demonstrates that the binary basicity had a prominent effect on the smelting effect. Fe and Cu 7 and demonstrates that the binary basicity had a prominent effect on the smelting effect. Fe and Cu recovery first rose and then dropped as the binary basicity increased from 0.3 to 1.5. The recovery recovery first rose and then dropped as the binary basicity increased from 0.3 to 1.5. The recovery of of Fe and Cu reached a maximum value of 98.6% and 97.2%, respectively, at 1.0 basicity when the Fe and Cu reached a maximum value of 98.6% and 97.2%, respectively, at 1.0 basicity when the basicity changed from 0.3 to 1.5. While the S content of the Fe–Cu master alloy decreased from 0.150% basicity changed from 0.3 to 1.5. While the S content of the Fe–Cu master alloy decreased from 0.150% to 0.039% during the binary basicity test. These results suggest that approximate basicity is in favour to 0.039% during the binary basicity test. These results suggest that approximate basicity is in favour of the separation between the Fe–Cu alloy and slag, and the removal of S. of the separation between the Fe–Cu alloy and slag, and the removal of S. The influence of binary basicity on the smelting process may be due to changes in slag viscosity. The influence of binary basicity on the smelting process may be due to changes in slag viscosity. The contour map of the change in the viscosity at various temperatures and basicity’s is shown in The contour map of the change in the viscosity at various temperatures and basicity’s is shown in Figure8. The map illustrates that raising the temperature or approximately increasing the binary Figure 8. The map illustrates that raising the temperature or approximately increasing the binary basicity is beneficial in improving the slag fluidity. It is worth noting that excessively high or low basicity is beneficial in improving the slag fluidity. It is worth noting that excessively high or low binary basicity results in a higher viscosity of the molten slag. Limestone can break the silicate and binary basicity results in a higher viscosity of the molten slag. Limestone can break the silicate and aluminosilicate bonds, and thereby increase the fluidity of the slag [24]. However, when the basicity aluminosilicate bonds, and thereby increase the fluidity of the slag [24]. However, when the basicity is over 1.1 (e.g., the addition of limestone is excessive), less liquid phase (as seen in Figure9) and is over 1.1 (e.g., the addition of limestone is excessive), less liquid phase (as seen in Figure 9) and more refractory solids, such as Ca2SiO4 (as seen in Figure 3), are generated, therefore, resulting in a molten slag with poor fluidity, which is detrimental to the smelting kinetics and phase separation.

Metals 2019, 9, 701 7 of 13

Metals 7 of 18 Metalsmore refractory solids, such as Ca2SiO4 (as seen in Figure3), are generated, therefore, resulting7 inof 18 a molten slag with poor ﬂuidity, which is detrimental to the smelting kinetics and phase separation. Ultimately, a basicity over 1.1 leads to low recovery of the Fe and Cu. Moreover, the viscosity of the Ultimately, aa basicitybasicity over over 1.1 1.1 leads leads to to low low recovery recovery of of the the Fe Fe and and Cu. Cu. Moreover, Moreover, the the viscosity viscosity of the of slagthe slag will tend to be at its minimum when the basicity is controlled in a range of 0.9~1.1. Therefore, willslag tendwill totend be to at itsbe minimumat its minimum when thewhen basicity the basicity is controlled is controlled in a range in ofa range 0.9~1.1. of Therefore,0.9~1.1. Therefore, based on based on the experimental results, the optimum basicity is recommended at 1.0. basedthe experimental on the experimental results, the results, optimum the optimum basicity is basicity recommended is recommended at 1.0. at 1.0.

99 100 0.30 99 100 0.30 98 98 98 Fe 98 0.25 Fe 97 TFe 0.25 TFe 96 97 96 96 0.20 96 94 0.20 Cu 94 95 Cu 95 92 92 0.15 94 0.15 94 90 90 1.8 content/% S

1.8 content/% S 93 TS 1.6 0.10

93 TS TCu Fe Cu and grade/% 1.6 0.10

TCu Fe Cu and grade/% 1.4 Fe Fe Cu and recovery/% 1.4 Fe Fe Cu and recovery/% 92 1.2 92 1.2 0.05 1.0 0.05 91 Fe recovry Fe grade 1.0 91 S content of alloy 0.8 Fe recovry Cu recovery Fe grade Cu grade S content of alloy 0.8 Cu recovery Cu grade 0.6 90 0.6 0.00 90 0.3 0.6 0.9 1.2 1.5 0.00 0.3 0.6 0.9 1.2 1.5 Binary basicity Binary basicity Figure 7. Effect of basicity on Fe and Cu recovery and S removal (Smelting at 1550 °C for 40 min). Figure 7. EffectEﬀect of basicity on Fe and C Cuu recovery and S removal (Smelting at 1550 ◦°C for 40 min).min).

1.6 1.6 Pa·S 1.700 Pa·S 1.700 1.300 0.2800 1.300 0.2800 1.4 0.9000 0.4500 1.4 0.9000 0.4500 0.4500 0.9000 0.4500 0.9000 1.2 0.2800 1.300 1.2 0.2800 1.300 1.700 1.700 1.0 2.100 1.0 2.100 2.500 2.500 0.8 2.900 0.8 0.2800 2.900 0.2800 3.200 3.200 Binary basicity Binary 0.6 0.4500 Binary basicity Binary 0.6 0.4500 0.9000 0.9000 0.4 1.300 0.4 1.300 2.100 1.700 2.100 1.700 0.2 2.500 0.2 2.500 1525 1550 1575 1600 1525 1550 1575 1600 o Temperature/o C Temperature/ C FigureFigure 8. 8.Contour Contour map map of of viscosity viscosity change change with with various various temperatures temperatures and and basicity. basicity. Figure 8. Contour map of viscosity change with various temperatures and basicity.

Metals 2019, 9, 701 8 of 13 Metals 8 of 18

1600 29.40 92.93 92.93 38.48

47.55 1580 83.85 56.63

74.78 65.70 1560 74.78

65.70 83.85

C o 92.93 T/ 56.63 1540 102.0 47.55

38.48 1520

1500 0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7 Basicity Figure 9. EffectEﬀect of basicity and temperature on percentage of liquid phase in slag.slag.

3.3. Mechanism of Desulfurization 3.3. Mechanism of Desulfurization It is well known that sulfur has adverse effects effects on the performance of steel products, such as its strength, ductility, ductility, and and toughness toughness [25 [25].]. As As noted noted in inthe the above above experimental experimental results, results, the thebasicity basicity of the of theslag slag plays plays a significant a significant role role in in desulfurization desulfurization during during the the smelting smelting process. process. To To further further reveal reveal the mechanismmechanism of desulfurization, the sulfidesulfide capacity of the slag was calculated and the desulfurization kinetics were investigated. 3.3.1. Sulfide Capacity (CS) of Slag 3.3.1. Sulfide Capacity (CS) of Slag The sulfide capacity (CS) and equilibrium distribution ratio of sulfur (LS) are important indicators The sulfide capacity (CS) and equilibrium distribution ratio of sulfur (LS) are important for measuring the desulfurization capacity of a slag system. In this paper, the optical basicity model indicators for measuring the desulfurization capacity of a slag system. In this paper, the optical was introduced to predict the CS of various slags with different basicity’s [23,26]. Optical basicity was basicity model was introduced to predict the CS of various slags with different basicity’s [23,26]. calculated by the model shown in Equation (2): Optical basicity was calculated by the model shown in Equation (2): Xn푛 ΛΛ== ∑ χ휒퐵BΛΛ퐵B (2(2)) B퐵==1 Where Λ is optical basicity, 휒 is mole fraction of oxide cations, which is also defined as the where Λ is optical basicity, χB is mole퐵 fraction of oxide cations, which is also defined as the fraction of eachfraction cationic of each charge cationic neutralizing charge neutralizing the negative the charge. negative charge. Hence, the 휒 can be calculated as the following: Hence, the χB퐵 can be calculated as the following: ‘ ‘ 휒퐵 = 푛표휒퐵/ X∑ 푛표휒퐵 (3) = 0 0 χB noχB/ noχB (3) ‘ Where 휒퐵 is the mole fraction of oxides, 푛표is the number of oxygen in each oxide component. 0 whereCombiningχB is the mole Equation fraction (2) ofwith oxides, (3), Young’sno is the model number was of oxygenadopted in to each calculate oxide the component. sulfide capacity in EquationCombining (4), because Equation it (2)is an with easy (3), calculation, Young’s model has a was wide adopted application to calculate scope theand sulfide a high capacitylevel of inaccuracy Equation [27,28]. (4), because it is an easy calculation, has a wide application scope and a high level of accuracy [27,28]. 11710 lg퐶푆 = −13.913 + 42.84훬 − − 0.02223푤(SiO2) − 0.02275푤(Al2O3) (4) 푇 11710 lgCS = 13.913 + 42.84Λ 0.02223w(SiO2) 0.02275w(Al2O3) (4) Where CS is the sulfide− capacity of slag,− TΛ is −optical basicity; 푤−(푆푖푂2) is mass fraction of SiO2 in slag, 푤(퐴푙2푂3) is mass fraction of Al2O3 in slag. ( ) whereBasedCS is on the Equation sulfide capacity (4), the sulfide of slag, capacityΛ is optical of the basicity; variousw SiOslags2 withis mass different fraction basicity’s of SiO2 incan slag, be ( ) wcalculatedAl2O3 is and mass the fraction results are of Al shown2O3 in in slag. Figure 10. The temperature and basicity have a positive effect on the sulfide capacity of the slag. The sulfide capacity increases by increasing the basicity and temperature. This result means that the slag with the higher basicity has a greater ability to hold and

Metals 2019, 9, 701 9 of 13

Based on Equation (4), the sulfide capacity of the various slags with different basicity’s can be calculated and the results are shown in Figure 10. The temperature and basicity have a positive Metalseffect on the sulfide capacity of the slag. The sulfide capacity increases by increasing the basicity9 of and 18 temperature. This result means that the slag with the higher basicity has a greater ability to hold and remove more sulfide sulfide from the metal phase, resulting in an an improvement improvement in in desulfur desulfurizationization efficiency efficiency and lower SS contentcontent inin thethe Fe–Cu Fe–Cu master master alloy. alloy. The The results results of of the the theoretical theoretical calculations calculations agree agree with with the theexperiments experiments shown shown in Figures in Figures5 and 5 7and. 7.

4.5 1525C 4.0 1550C 3.5 1575C 1600C 3.0

2.5

-4

/10 2.0 CS 1.5

1.0

0.5

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 Binary basicity Figure 10. EffectEffect of basicity and temperature on CS of the slag.slag. 3.3.2. Desulfurization Reaction Kinetics of Slag with Different Basicity 3.3.2. Desulfurization Reaction Kinetics of Slag with Different Basicity As previous work has shown, the desulfurization process consists of five steps: (1) mass transfer As previous work has shown, the desulfurization process consists of five steps: (1) mass transfer of [S] from the metal phase to the interface between the metal and slag; (2) mass transfer of (O2-) from of [S] from the metal phase to the interface between the metal and slag; (2) mass transfer of (O2-) from the molten slag to the interface between the metal and slag; (3) an interfacial electrochemical reaction at the molten slag to the interface between the metal and slag; (3) an interfacial electrochemical reaction the interface of the metal and slag; (4) mass transfer of the S2- generated in interfacial chemical reaction at the interface of the metal and slag; (4) mass transfer of the S2- generated in interfacial chemical from the interface to the molten slag; (5) mass transfer of O2- generated in the interfacial chemical reaction from the interface to the molten slag; (5) mass transfer of O2- generated in the interfacial reaction from the interface to the molten metal [28,29]. chemical reaction from the interface to the molten metal [28,29]. Compared with the diffusion steps, the rate of the electrochemical interfacial reaction is much Compared with the diffusion steps, the rate of the electrochemical interfacial reaction is much faster under a high smelting temperature. In view of that, the overall desulfurization rate is controlled faster under a high smelting temperature. In view of that, the overall desulfurization rate is controlled by the mass transfer. Moreover, the diffusion coefficient of S is tens of times higher than that of by the mass transfer. Moreover, the diffusion coefficient of S is tens of times higher than that of O, O, and the diffusion coefficient of sulfur in the metal is about two orders of magnitude larger than and the diffusion coefficient of sulfur in the metal is about two orders of magnitude larger than that that in the slag phase [28–30]. Hence, the control process of the desulfurization reaction is the mass in the slag phase [28–30]. Hence, the control process of the desulfurization reaction is the mass transfer of S in the slag, which was also confirmed by previous work [28,30–32]. Based on this result, transfer of S in the slag, which was also confirmed by previous work [28,30–32]. Based on this result, the desulfurization rate equation can be expressed as the following: the desulfurization rate equation can be expressed as the following: 푑d[[%S%S]] 퐴A − == ‧퐾.K{[[%S%S]] −[[%%S푆]]∗∗ (5(5)) − 푑푡dt 푉V푚m { − ∗ whereW [%S]here is[%S] the is S the content S content of molten of molten metal, metal,[%S]∗ [is% the푆] is S contentthe S content of the of metal-slag the metal interface,-slag interface, K is the K isdi fftheusion diffusion coefficient, coefficient,Vm is V them volumeis the volume of molten of molten metal, metal,A is the A interfaceis the interface area. area. By integrating Equation (5), the following rate formulas could be deduced as follows:

[%S]0 − [%S]퐹 1 퐼푛 [%S] [%S]F = 푘1 푡 (6) In[%S]0 − [%S] = k 퐻t (6) [%S]푡 [%S] 퐹 H t − F where [%S]0 is the S content in the molten metal at the initial stage, [%S]t is the S content in the molten metal at a time (t) during the reaction, [%S]F is the S content in the molten metal under equilibrium conditions, H is the height of the molten metal, k is the diffusion coefficient, t is the reaction time. Figure 11 shows the change in sulfur content with time and slag basicity in the alloy samples. By prolonging the reaction time, the S content in the metal decreased. Additionally, increasing the slag basicity can also significantly lower the S content in the metal.

Metals 2019, 9, 701 10 of 13

where [%S]0 is the S content in the molten metal at the initial stage, [%S]t is the S content in the molten metal at a time (t) during the reaction, [%S]F is the S content in the molten metal under equilibrium conditions, H is the height of the molten metal, k is the diﬀusion coeﬃcient, t is the reaction time. Figure 11 shows the change in sulfur content with time and slag basicity in the alloy samples. Metals 10 of 18 By prolonging the reaction time, the S content in the metal decreased. Additionally, increasing the slag

Metalsbasicity can also signiﬁcantly lower the S content in the metal. 10 of 18 0.25 0.7 basicity of slag 1.0 basicity of slag 1.6 basicity of slag 0.25 0.7 basicity of slag 0.20 1.0 basicity of slag 1.6 basicity of slag 0.20 0.15

0.15

0.10

S content of metal/% of content S 0.10 0.05

S content of metal/% of content S 0.05 0.00

0.00 0 20 40 60 80 100 120 Time/min Figure 11. Effects of time0 and basicity20 on40 S content60 of metal80 (Smelting100 temperature120 1550 °C). Time/min Figure 11. Eff[ects%S] of−[%S time] and basicity on S content of metal (Smelting temperature 1550 C). The Figure plots 11. of Effects퐼푛 0of time퐹 andwith basicity the reaction on S content times of ofmetal for (Smelting different temperature slag basicity 1550 are ◦°C) shown. in [%S]푡−[%S]퐹 [%S]0 [%S]F [%S] −[%S] [ −] [ ] 0 퐹 FigureThe 12, plots indicating of In [%S %Sthat] [0%S− the%S] with 퐹test thedates reaction (퐼푛 times of) for consist different well slag with basicity the theoretical are shown curve in Figure in the 12, The plots of 퐼푛 t F with the reaction[%S]푡−[ %S times]퐹 of for different slag basicity are shown in [%S−] −[%S] 푡 퐹[%S]0 [%S]F smeltingindicating process that the at test the dates various (In basicity’s− ) consistof[%S slag.]0− well[ %SThe]퐹 with standard the theoretical deviation curve value in theof the smelting fitted processcurves [%S]t [%S]F Figure 12, indicating that the test dates− (퐼푛 [%S]0−[%S) ]consist퐹 well with the theoretical curve in the wasat the all various over basicity’s 0.96, which of slag. revealed The standard that [퐼푛%S] deviation푡−[%S]퐹 value~t was of the a preferable fitted curves linear was relationship.all over 0.96, [ ] [ ] smelting process at the[%S various] [%S] basicity’s of slag.%S 푡The− %S standard퐹 deviation value of the fitted curves which revealed that In 0− F ~t was a preferable linear relationship. Furthermore, it confirmed that Furthermore, it confirmed[%S] that[%S] the control process[%S]0 −of[%S the]퐹 total desulfurization reaction was the mass was all over 0.96, whicht− revealedF that 퐼푛 ~t was a preferable linear relationship. transferthe control of [S] process in the of metal the total phase. desulfurization reaction[%S]푡−[%S was]퐹 the mass transfer of [S] in the metal phase. Furthermore, it confirmed that the control process of the total desulfurization reaction was the mass transfer of [S] in the metal phase. R 0.7 6 R 1.0 R 1.6 2 RFitted 0.7 curve R =0.98 2

} 65 RFitted 1.0 curve R =0.96 F RFitted 1.6 curve R2=0.97 Fitted curve R2=0.98

-[%]S }

t 5 Fitted curve R =0.96 F 4

Fitted curve R2=0.97

[%]S

(

-[%]S t ) 4

F 3

[%]S

(

-[%]S

) 0

F 32

[%]S

( -[%]S 0 2

In{ 1

[%]S (

In{ 10 0 20 40 60 80 100 120 Time/min 0 0 20 40 60 80[%S]0−[%S100]F 120 Figure 12. Relation curve between 퐼In 0− 퐹 and t. Figure 12. Relation curve betweenTime/min [%S]t [%S]F and t. [%S]푡−[%S]퐹 According to the slope of the line for the fitted curves[%S ] (as−[%S seen] in Figure 12), the diffusion According to the slopeFigure of the 12 line. Relation for the curve fitted between curves (as퐼n seen0 in Figure퐹 and t12. ), the diffusion coefficient coefficient can be determined, and the results are shown [ in%S ] Table푡−[%S]퐹 3. The diffusion coefficient for can be determined, and the results are shown in Table3. The di ffusion coefficient for desulfurization According to the slope of the line for the fitted curves−3 (as seen in Figure 12), the diffusion desulfurization reaction of 0.7 basicity slag was3 0.78 × 10 cm/s, whereas increasing the basicity to 1.0 reaction of 0.7 basicity slag was 0.78 10− cm/s, whereas increasing the basicity to 1.0 and 1.6, coefficientand 1.6, the can diffusion be determined, coefficient and elevated the× results to 0.92 are × 10 shown−3 cm/s inand Table 1.05 3.× 10 The−3 cm/s, diffusion respectively. coefficient These for desulfurizationresults mean that reaction the slag of basicity 0.7 basicity can promoteslag was the0.78 mass × 10− 3transfer cm/s, whereas of S in the increasing slag phase the and basicity accelerate to 1.0 andthe progress1.6, the diffusion of the desulfurizationcoefficient elevated reaction. to 0.92 As × 10 a− 3 result,cm/s and the 1.05 higher × 10 − basicity3 cm/s, respectively. allows for These better resultsdesulfurization mean that efficiency the slag basicityand lower can S promote content ofthe the mass Fe– transferCu alloy. of S in the slag phase and accelerate the progress of the desulfurization reaction. As a result, the higher basicity allows for better desulfurization efficiencyTable and 3. lowerDiffusion S content coefficient of theof sulphur Fe–Cu inalloy. alloy melt (cm/s).

Table 3.Items Diffusion coefficient of sulphurBasicity in alloy R melt (cm/s).

Items Basicity R

Metals 2019, 9, 701 11 of 13 the diffusion coefficient elevated to 0.92 10 3 cm/s and 1.05 10 3 cm/s, respectively. These results × − × − mean that the slag basicity can promote the mass transfer of S in the slag phase and accelerate the progress of the desulfurization reaction. As a result, the higher basicity allows for better desulfurization efficiency and lower S content of the Fe–Cu alloy.

Table 3. Diﬀusion coeﬃcient of sulphur in alloy melt (cm/s).

Metals Basicity R 11 of 18 Items 0.7 1.0 1.6 0.73 1.0 3 1.6 3 Diﬀusion coeﬃcient 0.78 10− 0.92 10− 1.05 10− Diffusion coefficient ×0.78 × 10−3 0.92 × 10−3 1.05 × 10−×3

3.4. Characterization of the Fe Fe–Cu–Cu Master Alloy The superior Fe–CuFe–Cu master alloy can be obtained under the optimum smelting conditions of 1550 °C,◦C, 40 min min and and 1.0 1.0 basicity. basicity. The The chemical chemical composition composition of ofthe the Fe Fe–Cu–Cu master master alloy alloy is shown is shown in inTable Table 4.4 The. The alloy alloy contains contains 95.9% 95.9% Fe, Fe, 1.4% 1.4% Cu Cu and and some minor impurities. impurities. The total total content content of of valuable metals is as high as 97.3%. In In addition, addition, the the S S content content of of the the master master alloy alloy decreased decreased to to 0.055%. 0.055%. The ﬁnalfinal product can be used along with scrap steel for the preparation of Cu Cu-bearing-bearing steel by electric arc furnace. Table 4. Chemical compositions of Fe–Cu master alloy (wt %). Table 4. Chemical compositions of Fe–Cu master alloy (wt %).

Fe Cu S SiO2 Al2O3 CaO MgO Na2O P Zn Pb Fe 95.86Cu 1.36 0.055S SiO 0.102 Al 0.322O3 0.053CaO 0.027MgO 0.005Na2O 0.01P 0.01Zn 0.009Pb 95.86 1.36 0.055 0.10 0.32 0.053 0.027 0.005 0.01 0.01 0.009 Figure 13 shows the microstructures of master alloy and chemical compositions under SEM and EDS. These results show that the master alloy is very uniform and pure. Through EDS analysis of a micro zone, the alloy contains 98.6% Fe and 1.4% Cu, and no obvious impurities were found, which further conﬁrmsconfirms the good performance of this master alloy.alloy.

Figure 13. SEMSEM-EDS-EDS and and microstructure microstructure of Fe Fe–Cu–Cu alloy alloy..

Metals 2019, 9, 701 12 of 13

4. Conclusions Experimental studies for the smelting process of CBDRI to prepare the Cu-bearing steel master alloy have been carried out in this work, and the mechanism of desulfurization was revealed as well. The following conclusions can be drawn: (1) A high quality Fe–Cu master alloy, assaying 95.9% Fe, 1.4% Cu and minor impurities, can be obtained from the smelting process under the optimum conditions of 1550 ◦C, 40 min and 1.0 basicity. The corresponding iron and copper recoveries are 98.6% and 97.2%, respectively. Through the smelting process, the total metal content in the alloy was increased, and the S content was significantly decreased to 0.05%, which is beneficial to the production of clean steel. It also confirms the necessity of further smelting for CBDRI. (2) Suitable basicity of the molten slag favours an increase in the recovery of Fe and Cu, which results from a rise in the amount of liquid and the improvement in the slag fluidity. In particular, the optimum binary basicity is in the range of 0.9~1.1. (3) The theoretical calculation results show that the sulfur capacity of the slag elevated with the increasing temperature and basicity, which benefits the desulfurization efficiency. The desulfurization reaction kinetics show that increasing the slag basicity contributes to the diffusion coefficient of the sulfur in the molten slag.

Author Contributions: D.Z., Z.G. and J.P. conceived and designed the experiments; Z.G. and L.P. performed the experiments; Z.G. and L.P. wrote the paper; D.Z. modified the paper. Funding: This research was funded by National Natural Science Foundation of China (No.51474161) and Innovation-driven Project of Guangxi Zhuang Autonomous Region (No. AA18242003, No. AA148242003). Acknowledgments: The authors wish to express thanks to the National Natural Science Foundation of China (No. 51474161) for the financial support of this research, and also would like to thank Co-Innovation Center for Clean and Efficient Utilization of Strategic Metal Mineral Resources of Hunan Province, which supplied us the facilities and funds to fulfill the experiments. Conflicts of Interest: The authors declare no conflict of interest.

References

1. Shi, X.B.; Yan, W.; Xu, D.K.; Yan, M.C.; Yang, C.G.; Shan, Y.Y.; Yang, K. Microbial corrosion resistance of a novel Cu-bearing pipeline steel. J. Mater. Sci. Technol. 2018, 34, 2480–2491. [CrossRef] 2. Cao, Z.Q.; Zhao, J.; Yang, K. Cu-bearing stainless steel reduces cytotoxicity and crystals adhesion after ureteral epithelial cells exposing to calcium oxalate monohydrate. Sci. Rep. 2018, 8.[CrossRef][PubMed] 3. Zhang, X.R.; Zhao, J.L.; Xi, T.; Shahzad, M.B.; Yang, C.G.; Yang, K. Dissolution and repair of passive ﬁlm on

Cu-bearing 304L stainless steels immersed in H2SO4 solution. J. Mater. Sci. Technol. 2018, 34, 2149–2159. [CrossRef] 4. Morcillo, M.; Díaz, I.; Chico, B.; Cano, H.; Fuente, D. Weathering steels: From empirical development to scientiﬁc design. A review. Corros. Sci. 2014, 83, 6–31. [CrossRef] 5. Morcillo, M.; Chico, B.; Díaz, I.; Cano, H.; Fuente, D. Atmospheric corrosion data of weathering steels. A review. Corros. Sci. 2013, 77, 6–24. [CrossRef] 6. Ma, T.; Yang, G.Y.; Deng, M.L.; Liu, Y.G. Research Status and Prospect of Copper-bearing Steel. Hot Work. Technol. 2017, 46, 36–39. (In Chinese) 7. Shyi, W.; Powe, K. Eﬀect of alloying elements on the structure and mechanical properties of ultra low carbon bainitic steels. Mater. Sci. 1993, 28, 5169–5175. 8. Guo, Z.Q.; Pan, J.; Zhu, D.Q.; Zhang, F. Co-reduction of Copper Smelting Slag and Nickel Laterite to Prepare Fe-Ni-Cu Alloy for Weathering Steel. JOM 2018, 70, 150–154. [CrossRef] 9. 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]

10. Guo, Z.Q.; Zhu, D.Q.; Pan, J.; Yao, W.J.; Xu, W.Q.; Chen, J.A. Eﬀect of Na2CO3 Addition on Carbothermic Reduction of Copper Smelting Slag to Prepare Crude Fe-Cu Alloy. JOM 2017, 69, 1688–1695. [CrossRef] Metals 2019, 9, 701 13 of 13

11. Gorai, B.; Jana, R.; Premchand, M. Characteristics and utilization of copper slag-a review. Resour. Conserv. Recycl. 2003, 39, 299–313. [CrossRef] 12. Subrata, R.; Amlan, D.; Sandeep, R. Flotation of copper sulphide from copper smelter slag using multiple collectors and their mixtures. Int. J. Miner. Process. 2015, 34, 43–49. 13. Guo, Z.Q.; Pan, J.; Zhu, D.Q.; Zhang, F. Green and eﬃcient utilization of waste ferric-oxide desulfurizer to clean waste copper slag by the smelting reduction-sulfurizing process. J. Clean. Prod. 2018, 199, 891–899. [CrossRef] 14. Li, Y.; Perederiy, I.; Papangelakis, V.G. Cleaning of waste smelter slags and recovery of valuable metals by pressure oxidative leaching. J. Hazard. Mater. 2008, 152, 607–615. [CrossRef][PubMed] 15. Shibayama, A.; Takasaki, Y.; William, T.; Yamatodani, A.; Higuchi, Y.; Sunagawa, S.; Ono, E. Treatment of smelting residue for arsenic removal and recovery of copper using pyro–hydrometallurgical process. J. Hazard. Mater. 2010, 181, 1016–1023. [CrossRef][PubMed] 16. Perederiy, I.; Papangelakis, V.G.; Buarzaiga, M.; Mihaylov, I. Co-treatment of converter slag and pyrrhotite tailings via high pressure oxidative leaching. J. Hazard. Mater. 2011, 194, 399–406. [CrossRef] 17. Antonijevic, M.M.; Dimitrijevic, M.D.; Stevanovic, Z.O.; Serbula, S.M.; Bogdanovic, G.D. Investigation of the possibility of copper recovery from the ﬂotation tailings by acid leaching. J. Hazard. Mater. 2008, 152, 23–34. [CrossRef] 18. Zuo, Z.L.; Yu, Q.B.; Xie, H.Q.; Wang, K.; Liu, S.H.; Yang, F.; Qin, Q.; Qi, Z.F. Mechanical and reduction characteristics of cold-pressed copper slag pellets composited within biomass and lignite. Renew. Energy 2018, 125, 206–224. [CrossRef] 19. 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] 20. Chun, T.; Mu, G.; Di, Z.; Long, H.; Ning, C.; Li, D. Recovery of iron from copper slag by carbothermic thermic reduction and magnetic separation in the presence of CaO. Arch. Metall. Mater. 2018, 63, 299–305. 21. Geng, C.; Wang, H.J.; Hu, W.T.; Li, L.; Shi, C.S. Recovery of iron and copper from copper tailings by coal-based direct reduction and magnetic separation. J. Iron Steel Res. Int. 2017, 24, 991–997. [CrossRef] 22. Jiao, R.M.; Xing, P.; Wang, C.Y.; Ma, B.Z.; Chen, Y.Q. Recovery of iron from copper tailings via low-temperature direct reduction and magnetic separation: Process optimization and mineralogical study. Int. J. Min. Met. Mater. 2017, 24, 974–982. [CrossRef] 23. Wang, X.H. Ferrous Metallurgy (Steelmaking Part), 1st ed.; Metallurgical Industry Press: Beijing, China, 2007. (In Chinese) 24. Liu, Y.; Lv, X.; Xu, J.; Zhang, S.; Bai, C. Preparation of stainless steel master alloy by direct smelting reduction

of Fe–Ni–Cr Sinter at 1600 ◦C. Ironmak. Steelmak. 2016, 8, 1–7. [CrossRef] 25. Yan, Z.; Lv, X.; Pang, Z.; He, W.; Liang, D.; Bai, C. Transition of Blast Furnace Slag from Silicate Based to Aluminate Based: Sulfide Capacity. Metall. Mater. Trans. B 2017, 48, 2607–2613. [CrossRef] 26. Guo, H. Metallurgical Physical Chemistry, 2nd ed.; China Metallurgical Industry Press: Beijing, China, 2012. (In Chinese) 27. Young, R.W.; Duffy, J.A.; Hassall, G.J. Use of optical basicity concept of determining phoshphorus and sulfur slag-metal partitions. Ironmak. Steelmak. 1992, 19, 201–209. 28. Wang, L.; Wu, S.; Kou, M.; Du, B.; Lu, Y.; Gu, K. Improving the Desulphurization in COREX Process by Adjusting the Hot Metal Chemical Composition. Metall. Mater. Trans. B 2018, 49, 89–96. [CrossRef] 29. Wang, X.L. Ferrous Metallurgy (Ironmaking Part), 3rd ed.; China Metallurgical Industry Press: Beijing, China, 2013. (In Chinese) 30. Kang, J.G.; Shin, J.H.; Chung, Y.; Park, J.H. Effect of Slag Chemistry on the Desulfurization Kinetics in Secondary Refining Processes. Metall. Mater. Trans. B 2017, 48, 2123–2135. [CrossRef] 31. Peter, J.; Peaslee, K.D.; Robertson, D.G.C.; Thomas, B.G. Proceedings of the AISTech 2005; AIST: Warrendale, PA, USA, 2005; pp. 959–967. 32. Robertson, D.G.C.; Staples, B.B. Process Engineering of Pyrometallurgy; IMM: London, UK, 1974; pp. 51–59.

© 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/).