Extraction of Cobalt and Iron from Refractory Co-Bearing Sulfur Concentrate

Authors:

Junhui Xiao, Yushu Zhang

Date Submitted: 2020-04-14

Keywords: roasting, magnetic separation, Co-bearing sulfur concentrate, iron, cobalt

Abstract:

In this study, oxidizing roasting, segregation roasting, and magnetic separation are used to extract cobalt and iron from refractory Co- bearing sulfur concentrate. The Co-bearing sulfur concentrate containing 0.68% Co, 33.26% Fe, and 36.58% S was obtained from V-Ti magnetite in the Panxi area of China by flotation. Cobalt pyrite and linneite were the Co-bearing minerals, and the gangue minerals were mica, chlorite, feldspar, and calcite in Co-bearing sulfur concentrate. The results show that cobalt is transformed from Co-pyrite and linneite to a Co2FeO4-dominated new cobalt mineral phase, and iron is transformed from pyrite to Fe2O3 and an Fe3O4-dominated new iron mineral phase after oxidizing roasting. Cobalt changed from CoFe2O4 to a new cobalt mineral phase dominated by [Co] Fe solid solution, and iron changed from Fe2O3 to a new iron mineral phase dominated by metal Fe and Fe3O4 after segregation roasting. Cobalt concentrate with a cobalt grade of 15.15%, iron content of 71.22%, and cobalt recovery of 90.81% as well as iron concentrate with iron grade of 60.06%, cobalt content of 0.11%, and iron recovery of 76.23% are obtained. The main minerals in the cobalt concentrate are Fe, [Co]Fe, Fe3O4, and SiO2, and the main minerals in the iron concentrate are Fe3O4, FeO, Ca2Si2O4, and Ca2Al2O4.

Record Type: Published Article

Submitted To: LAPSE (Living Archive for Process Systems Engineering)

Citation (overall record, always the latest version): LAPSE:2020.0345 Citation (this specific file, latest version): LAPSE:2020.0345-1 Citation (this specific file, this version): LAPSE:2020.0345-1v1

DOI of Published Version: https://doi.org/10.3390/pr8020200

License: Creative Commons Attribution 4.0 International (CC BY 4.0)

Article Extraction of Cobalt and Iron from Refractory Co-Bearing Sulfur Concentrate

Junhui Xiao 1,2,3,4,* and Yushu Zhang 2,4

1 Sichuan Engineering Lab of Non-Metallic Mineral Powder Modiﬁcation and High-Value Utilization, Southwest University of Science and Technology, Mianyang 621010, China 2 Institute of Multipurpose Utilization of Mineral Resources, Chinese Academy of Geological Sciences, Chengdu 610041, China; [email protected] 3 Key Laboratory of Sichuan Province for Comprehensive Utilization of Vanadium and Titanium Resources, Panzhihua University, Panzhihua 61700, China 4 Key Laboratory of Ministry of Education for Solid Waste Treatment and Resource Recycle, Southwest University of Science and Technology, Mianyang 621010, China * Correspondence: [email protected]; Tel.: +86-139-9019-0544

Received: 7 January 2020; Accepted: 4 February 2020; Published: 6 February 2020

Abstract: In this study, oxidizing roasting, segregation roasting, and magnetic separation are used to extract cobalt and iron from refractory Co-bearing sulfur concentrate. The Co-bearing sulfur concentrate containing 0.68% Co, 33.26% Fe, and 36.58% S was obtained from V-Ti magnetite in the Panxi area of China by ﬂotation. Cobalt pyrite and linneite were the Co-bearing minerals, and the gangue minerals were mica, chlorite, feldspar, and calcite in Co-bearing sulfur concentrate. The results show that cobalt is transformed from Co-pyrite and linneite to a Co2FeO4-dominated new cobalt mineral phase, and iron is transformed from pyrite to Fe2O3 and an Fe3O4-dominated new iron mineral phase after oxidizing roasting. Cobalt changed from CoFe2O4 to a new cobalt mineral phase dominated by [Co] Fe solid solution, and iron changed from Fe2O3 to a new iron mineral phase dominated by metal Fe and Fe3O4 after segregation roasting. Cobalt concentrate with a cobalt grade of 15.15%, iron content of 71.22%, and cobalt recovery of 90.81% as well as iron concentrate with iron grade of 60.06%, cobalt content of 0.11%, and iron recovery of 76.23% are obtained. The main minerals in the cobalt concentrate are Fe, [Co]Fe, Fe3O4, and SiO2, and the main minerals in the iron concentrate are Fe3O4, FeO, Ca2Si2O4, and Ca2Al2O4.

Keywords: cobalt; iron; Co-bearing sulfur concentrate; roasting; magnetic separation

1. Introduction The unique physical and chemical properties of cobalt make this metal an important raw material in the fields of aerospace, petrochemicals, glass manufacturing, and medicine. Cobalt plays an important role in the development of strategic emerging industries. The content of cobalt in the crust is very low, and 90% of cobalt in the crust is in a dispersed state. In nature, cobalt is mostly associated with copper, nickel, iron, and other minerals, and there is no separate cobalt deposit. Both the United States and the European Union have included cobalt in the list of key minerals and materials that affect national and regional security and future economic development. According to the research of the China Mining Federation, cobalt will become one of the nine minerals in serious shortage by 2020. China is one of the world’s major producers and consumers of refined cobalt, but China’s cobalt reserves only account for 1.1% of the total global reserves. A large number of cobalt raw materials rely on imports. Most of the cobalt minerals found on the earth are associated with copper sulfide ore, nickel sulfide ore, pyrite, and other minerals, and only about 2% of the global cobalt production

Processes 2020, 8, 200; doi:10.3390/pr8020200 www.mdpi.com/journal/processes Processes 2020, 8, 200 2 of 18 comes from independent cobalt mines. According to the statistics of USGS 2019, the proven global land cobalt resources are 25 million tons and reserves are 6.88 million tons [1]. Based on the global cobalt production of 136,000 tons in 2018, the static guarantee period is 50 years. More than 120 million tons of cobalt resources have been found at the bottom of the Atlantic Ocean, the Indian Ocean, and the Pacific Ocean. Because of technical and economic reasons, these resources have not been exploited up to now. The global land cobalt resources are widely distributed. Cobalt resources mainly occur in sedimentary stratiform copper cobalt deposits in the Congo and Zambia; laterite nickel cobalt deposits in Australia, Cuba, the Philippines, and Madagascar; and magmatic nickel copper sulfide deposits in Australia, Canada, and Russia. Although cobalt ore is widely distributed, except that the bouazzer cobalt ore in Morocco is a separate cobalt ore with arsenic cobalt as the main mineral, the rest of the world’s cobalt ore is produced as a co-associated mineral of copper, nickel, and other minerals. At present, only a few countries, such as the Congo, Australia, Cuba, Canada, and Russia, can use the cobalt ore economically. There are more than 100 kinds of minerals containing cobalt in nature, and more than 59 kinds of minerals contain cobalt as a basic element. However, at present, there are mainly three kinds of cobalt minerals with economic significance: arsenides, sulfides, and oxides. Common cobalt minerals in the industry include cobaltite, glaucodot, linnetie, smaltite, safflorite, and erthrite as well as asbolite and heterogenite in supergene minerals [2]. The main common cobalt minerals and their properties are shown in Table1.

Table 1. The common cobalt minerals and their properties.

Mineral Mineral Cobalt Density Moh’s Chemical Formula Species Names Content (%) (g cm 3) Hardness · − Smaltite CoAs2 1.5–24 6.4–6.6 5.6–6 Saﬄorite (Co, Ni, Fe)As2 1.3–23 6.9–7.3 4.5–5 Arsenide Skutterudite CoAs3(Co, Ni, Fe) As3 16–20 6.7–6.9 6 Cobaltite CoAsS 29–39 5.8–6.3 5.5–6 Glaucodot (Co, Fe)AsS 8–18 5.9–6 5 Carrollite Cu(Co, Ni) S 27–42 4.8–5 5.5 Sulﬁde 2 4 Linneite Co3S4 45–53 4.8–5 5.5 Asbolite CoO 2MnO 4H O Traces-25 3.1 2–2.5 · 2· 2 Erthrite Co (AsO ) 8H O 30 2.9–3 1.5–2.5 Oxide 3 4 2· 2 Heterogenite Co2O3H2O 50–60 4.1–4.5 3–5 Spherocobaltite CoCO3 50–55 4–4.1 4

Cobalt-sulfur concentrate is the most common raw material for industrial cobalt extraction. Pyrite is the most important sulfur-bearing mineral in cobalt sulfur concentrate. Because the cobalt-sulfur concentrate mainly contains pyrite, pyrrhotite, and some gangue minerals such as talcum, quartz, and chlorite, it is difficult to obtain separate cobalt and sulfur concentrates. Most cobalt minerals are associated with other metal minerals. At present, the main cobalt minerals to be processed include cobaltite, skutterudite, thiocopper cobalt, thiocobaltite, nickel cobalt, cobaltite, iron manganese combined cobalt, etc. The most common process to recover cobalt minerals is flotation, which also includes manual separation and gravity separation. Single cobalt minerals, such as cobaltite and thiocobaltite, can be directly obtained from cobalt concentrate by flotation; for cobalt existing in pyrite and chalcopyrite, carrier flotation is generally used, and cobalt concentrate cannot be directly obtained. Recovery of associated cobalt in V-Ti-magnetite is essentially a mixture of pyrite, pyrrhotite, cobaltite, and thiocobaltite. The recovery of valuable metals from cobalt-bearing sulfur concentrate is usually carried out by the combination of pyrometallurgy and hydrometallurgy [3–6]. For example, the main disadvantage of direct sulfuric acid roasting in the Shandong Zibo cobalt plant is that the treatment capacity is low (3.0 t/m2 d), and the SO content in the flue gas is only 4–5%, which is not conducive to · 2 acid production. Aiming at the roasting leaching of cobalt sulfur concentrate with high cobalt content, low sulfur content, and a considerable amount of copper, the effects of roasting temperature, roasting Processes 2020, 8, 200 3 of 18 time, additive dosage, leaching time, leaching temperature, and liquid–solid ratio were analyzed to investigate the influence of the leaching rates of copper and cobalt. After the cobalt-sulfur concentrate was mixed evenly, the temperature was raised to 620 ◦C at 2.7 ◦C/min, calcined for 3 h, and the calcine was leached at 80 ◦C with 33% slurry concentration of 30 g/L sulfuric acid for 2 h, the results showed leaching rates of Co and Cu of 91% and 90% respectively. The cobalt sulfur concentrate was oxidized roasted at 850–900 ◦C, and the cinder was mixed with some cobalt sulfur concentrate and chlorinating agent to carry out medium-temperature chlorination sulfuric acid roasting at about 650 ◦C. This way had a high bed capacity (17–18 t/m2 d), but the SO concentration in the flue gas was only 1–2%, which · 2 is difficult for acid production, and there was a certain amount of hydrogen chloride gas in the flue gas, which will lead to serious corrosion of the equipment [7–10]. There is a large amount of valuable mineral resources in the Panxi area of China, and the proven reserves of V-Ti magnetite are 10 billion tons. Among them, iron resources account for 20% of the domestic iron ore reserves, vanadium resources account for 62% of the national vanadium reserves, and titanium resources account for 90.5% of the national titanium reserves. In addition, there are 900,000 tons of cobalt, 700,000 tons of nickel, 250,000 tons of scandium, 180,000 tons of gallium, and other resources. The comprehensive utilization rate of nonferrous metal resources in the Panxi area of China is very low. Because of the low grade and scattered distribution, it is difficult to recover the sulfur and cobalt resources directly from the raw ore. Most of the associated valuable metal cobalt in V-Ti magnetite occurs in the form of pyrite, pyrrhotite, and cobaltite, and a very small amount of cobalt occurs in the form of thiocobaltite. Because the floatability difference among pyrite, pyrrhotite, cobalt nickel pyrite, and thiocobaltite is small, they basically exhibit the surface properties of pyrite minerals, so it is difficult to recover separate minerals by flotation. At present, iron concentrate, titanium concentrate, and cobalt-bearing sulfur concentrate are obtained from V-Ti magnetite ores in the Panxi area of China by magnetic separation and flotation. In the process of flotation, cobalt enters into sulfur concentrate, which is closely related to pyrite. Generally, Co-bearing sulfur concentrate with a cobalt grade of about 0.6% can be enriched by flotation. It is difficult to further separate cobalt and sulfur using flotation or other physical beneficiation. In this paper, a technology of two-stage roasting and magnetic separation is adopted to extract cobalt and iron from the complex Co-bearing sulfur concentrate in the Panxi area of China. This will provide an important research basis for the efficient utilization of Co-bearing sulfur concentrate resources in Panxi [11–14].

2. Materials and Methods

2.1. Sampling The Co-bearing sulfur concentrate sample was collected from a V-Ti magnetite dressing plant by ﬂotation in the Panzhihua area of China. The Co-bearing sulfur concentrate contained Co 0.68%, Fe 33.26%, and S 36.58%. Pyrite was the main sulﬁde ore. Co-pyrite and linneite were Co-bearing minerals. Gangue minerals were mica, chlorite, feldspar, calcite, etc., in the Co-bearing concentrate. The main chemical composition analysis of the sample is shown in Table2. The main mineral analysis of the sample is shown in Table3, a grain-sized analysis of the sample is shown in Table4, and the XRD analysis of the sample is shown in Figure1.

Table 2. Main chemical composition of Co-bearing sulfur concentrate (%).

Fe Co S CaO MgO Al2O3 SiO2 Mn K2O Na2O TiO2 33.26 0.68 36.58 5.06 3.67 8.63 8.26 0.06 0.68 0.23 2.06 Processes 2020, 8, 200 4 of 18 Processes 2020, 8, 200 4 of 18

Table 3. Main minerals composition of Co-bearing sulfur concentrate (%).

PyritePyrite Linneite Linneite Co-Pyrite Co-Pyrite Mica Mica Chlorite Chlorite Calcite Calcite Feldspar Feldspar Other Other Minerals Minerals 71.6571.65 3.68 3.68 2.65 2.65 2.64 2.64 3.36 3.36 8.22 8.22 4.68 4.68 3.12 3.12

Table 4. Grain-sized analysis of Co-bearing sulfur concentrate. Table 4. Grain-sized analysis of Co-bearing sulfur concentrate. Fractions Productivity Grade (%) Distribution Rate (%) Grade (%) Distribution Rate (%) (mm)Fractions (mm) (%)Productivity (%)Fe Co S Fe Co S −0.28~+0.154 10.68 30.27 Fe0.56 Co32.97 S Fe9.72 Co 8.80 S 9.63 0.28~+0.154 10.68 30.27 0.56 32.97 9.72 8.80 9.63 −0.154~+0.1− 24.15 34.59 0.72 37.86 25.12 25.57 25.00 0.154~+0.1 24.15 34.59 0.72 37.86 25.12 25.57 25.00 −0.1~+0.074− 36.78 33.18 0.68 37.22 36.69 36.78 37.42 0.1~+0.074 36.78 33.18 0.68 37.22 36.69 36.78 37.42 − −0.074~+0.0450.074~ +0.04517.25 17.25 33.69 33.690.65 0.65 35.84 35.84 17.4717.47 16.4916.49 16.90 16.90 − −0.045~+0.0380.045~ +0.038 6.36 6.36 34.02 34.020.78 0.7836.36 36.36 6.516.51 7.30 7.30 6.32 6.32 − −0.038+0 0.038+0 4.78 4.78 31.27 31.270.72 0.7236.21 36.21 4.494.49 5.06 5.06 4.73 4.73 − Totals Totals100.00 100.0033.26 33.260.68 0.6836.58 36.58 100.00100.00 100.00100.00 100.00 100.00

800 1-Pyrite

700 2-Linneite 3-Cobalt pyrite

600 4-Mica 5-Chlorite 6-Feldspar 500 1 7-Calcite

400 1

300 Intensity(a.u.) 1 2 3 7 200 5 1 2 6 3 6 7 100 4

0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 2θ (°) Figure 1. X-rayX-ray diffraction diffraction (XRD) diffractogram diffractogram of Co-bearing sulfur concentrate. 2.2. Chemical Reagent and Equipment 2.2. Chemical Reagent and Equipment The main chemical reagents used in this test were calcium chloride and sodium chloride with The main chemical reagents used in this test were calcium chloride and sodium chloride with analytical purity from Guangzhou Chemical Regent, Co., Ltd., Guangzhou China. The coke from analytical purity from Guangzhou Chemical Regent, Co., Ltd., Guangzhou China. The coke from Shanxi Coking Coal Group Co. Ltd. (Taiyuan, China) was crushed and divided into five particle sizes Shanxi Coking Coal Group Co. Ltd (Taiyuan, China). was crushed and divided into five particle sizes of 1 mm, 0.8 mm, 0.6 mm, 0.4 mm, and 0.2 mm as reducing agents in the process of segregation of −1 mm, −−0.8 mm, −−0.6 mm, −−0.4 mm, and −−0.2 mm as reducing agents in the process of segregation roasting. The entire quality analysis of coke is shown in Table5. roasting. The entire quality analysis of coke is shown in Table 5.

Table 5. The entire quality analysis of coke. Mad: Air drying base content. Ad: ash content. Vdaf: Table 5 The entire quality analysis of coke. Mad: Air drying base content. Ad: ash content. Vdaf: Volatile Volatile content. FCd: Fixed Carbon content. content. FCd: Fixed Carbon content.

Mad Ad FCd Vdaf Characteristic of Char Residue Mad Ad FCd Vdaf Characteristic of Char Residue 0.440.44 2.05 2.05 95.6895.68 1.081.08 2 2

The main equipment used in the experiment included an electric atmosphere tube furnace The main equipment used in the experiment included an electric atmosphere tube furnace (Model: (Model: SX-6-16, Changsha Kehui Furnace Technology Co., Ltd., Changsha, China), muffle furnace SX-6-16, Changsha Kehui Furnace Technology Co., Ltd., Changsha, China), muﬄe furnace ( 1300 C, (≤1300 °C, Shanghai Shiyan Electric Furnace Co., Ltd. Shanghai, China), cone ball mill (XMQ—≤ Ф 240◦ Shanghai Shiyan Electric Furnace Co., Ltd. Shanghai, China), cone ball mill (XMQ—Φ 240 90 mm, Jilin × 90 mm, Jilin Exploration Machinery Factory, Changchun, China), Disc grinder (Ф 300× × 150 mm, Exploration Machinery Factory, Changchun, China), Disc grinder (Φ 300 150 mm, Jilin Exploration Jilin Exploration Machinery Factory, Changchun, China), drying box (Shanghai× Shiyan Yan Electric Machinery Factory, Changchun, China), drying box (Shanghai Shiyan Yan Electric Furnace Co., Ltd. Furnace Co., Ltd. Shanghai China), and vacuum filter (Ф 300, Southwest Chengdu Experimental Shanghai China), and vacuum ﬁlter (Φ 300, Southwest Chengdu Experimental Equipment Co., Ltd. Equipment Co., Ltd. Chendu, China). Chendu, China).

Processes 2020, 8, 200 5 of 18

2.3. Oxidizing Roasting Test Oxidizing roasting tests were conducted in an electric atmosphere tube furnace, and Co-bearing sulfur concentrate was put into a corundum dry pot and put into the electric atmosphere tube furnace. The temperature was set to 600–1000 ◦C in an oxygen atmosphere. When the temperature of the electric atmosphere tube furnace rose to the set temperature, timing started. After roasting for 1–3 h, the oxygen supply was stopped, and the power supply was turned oﬀ. Meanwhile, the oxygen-passing tube was pulled out, and the roasting product was taken out for cooling, manual crushing, and grinding to <0.100 mm for segregation roasting. The oxidizing roasting conditions were conﬁrmed to be reasonable according to the analysis of cobalt, iron, and sulfur contents in oxidizing roasting ores. Calculation of the desulfurization rate is shown in Formula (1).

Desulfurization rate = (Q α Q β)/Q α 100% (1) 1 × − 2 × 1 × ×

Here, Q1 is the weight of Co-bearing sulfur concentrate, g; Q2 is the weight of oxidizing roasting ore, g; α is the sulfur content of Co-bearing sulfur concentrate, %; and β is the sulfur content of oxidizing roasting ore, %.

2.4. Segregation Roasting–Magnetic Separation Test The Co-bearing concentrate, the chlorinating agent, and reducing agent were mixed and put into the roasting furnace. These mixtures were heated to a certain temperature (400–1100 ◦C) under a neutral or weak reducing atmosphere, where the chlorinating agent was used to form volatile metal chlorides. For each magnetic test, segregation roasting ores were put into a 6.25 dm3Φ 240 90 conical × ball mill. The grinding density was set at 60%. The pulp was then placed in a XCGS-13 (Φ 50, Overall dimensions 1000 800 500 mm) Davis Magnetic Tube (Jilin Exploration Machinery Plant, Shenyang × × China) with special magnetic ﬁeld intensity. The products were ﬁltered, dried, weighed, sampled, and tested for further evaluation based on the grade and recovery of cobalt and iron in magnetic products. The calculation formulas of cobalt recovery and iron recovery are shown in Formula (2).

Cobalt recovery = Q β /Q α 100%, Iron recovery = Q β /Q α 100% (2) 3 × 1 0 × 1 × 4 × 2 0 × 2 ×

Here, Q3 is the weight of cobalt concentrate, g; Q4 is the weight of iron concentrate, g; Q0 is the weight of segregation roasting ores, g; β1 is the cobalt grade of cobalt concentrate, %; β2 is the iron grade of cobalt concentrate, %; α1 is the cobalt grade of segregation roasting ores, %; and α2 is the iron grade of segregation roasting ores, %.

2.5. Procedure The Co-bearing sulfur concentrate (mass of 100 g for each test) was placed in a tubular atmosphere furnace for oxidizing roasting. The reaction of the sulfide ore and oxygen was reversed to oxidize the ore, and the metals such as iron and cobalt were transformed into iron oxide and cobalt oxide. Oxidizing roasting ore (mass of 50 g for each test) after cooling, chlorinating agent, and coke were mixed and put into the muffle furnace for segregation chlorination roasting, which turned cobalt into a new metal cobalt mineral phase and iron into a new magnetite mineral phase. Metal cobalt is a ferromagnetic metal, and magnetite is a highly magnetic minerals. The specific magnetic susceptibility of cobalt metal is greater than the magnetic susceptibility of magnetite. Cobalt was recovered from segregation after grinding with a magnetic separator; iron was recovered from magnetic separation tailings by two-stage magnetic separation. The process of oxidizing roasting–segregation roasting–magnetic separation is shown in Figure2. Processes 2020, 8, 200 6 of 18 Processes 2020, 8, 200 6 of 18

Co-bearing sulfur concentrate Segregation roasting

Addition agent Mixing Cooling

Oxidizing roasting Grinding

Magnetic separation Cooling H1 Magnetic separation Coke Mixing and stoving (80℃) Cobalt concentrate H2

Chlorinating agent Iron concentrate Tailings

Figure 2. Scheme of the Co and Fe extraction process from Co-bearing sulfur concentrate.

2.6. AnalysisFigure and Characterization2. Scheme of the Co and Fe extraction process from Co-bearing sulfur concentrate. The chemical composition of solid materials was analyzed by a Z-2000 atomic absorption 2.6. Analysis and Characterization spectrophotometer (Hitachi Co., Ltd. Tokyo, Japan); the direction grating was Zenier-tana type, 1800 linesThe /chemicalmm; the flashcomposition wavelength of solid was materials 200 nm, the was wavelength analyzed rangeby a Z-2000 was 190~900 atomic nm; absorption and the automaticspectrophotometer peak seeking (Hitachi setting Co., and Ltd. the Tokyo, spectral Japan) bandwidth; the direction was divided grating into was 4 grades Zenier-tana (0.2, 0.4, type, 1.3, 1800 and 2.6lines/mm; nm) to analyzethe flash the wavelength chemical composition was 200 nm, of the th mineral.e wavelength range was 190~900 nm; and the automaticThe phase peak compositionsseeking setting of and solid the substances spectral bandwidth (Co-bearing was sulfur divided concentrate, into 4 grades oxidizing (0.2, roasting 0.4, 1.3, ores,and 2.6 segregation nm) to analyze roasting the ores, chemical cobalt composition concentrate, of iron the concentrate, mineral. and magnetic separation tailings) wereThe analyzed phase by compositions X-ray diffraction of solid (XRD, substances X Pert pro, (Co- Panaco,bearing Thesulfur Netherlands). concentrate, The oxidizing microstructure roasting ofores, the segregation solid products roasting was ores, observed cobalt concentrate, by SEM (S440, iron Hirschmann concentrate, and Laborgerate magnetic GmbH separation & Co. tailings) KG, Eberstadt,were analyzed Germany) by X-ray equipped diffraction with (XRD, an energy-dispersive X Pert pro, Panaco, X-ray The spectroscopy Netherlands). (EDS) The detector microstructure (UItra55, CarlzeissNTSof the solid products GmbH, Hirschmann was observed Laborgerate by SEM GmbH(S440, Hirschmann & Co. KG, Eberstadt, Laborgerate Germany). GmbH & Co. KG, Eberstadt, Germany) equipped with an energy-dispersive X-ray spectroscopy (EDS) detector 3.(UItra55, Results CarlzeissNTS and Discussion GmbH, Hirschmann Laborgerate GmbH & Co. KG, Eberstadt, Germany). 3.1. Removing Sulfur by Oxidizing Roasting 3. Results and Discussion 3.1.1. Effect of Roasting Temperature 3.1. Removing Sulfur by Oxidizing Roasting The roasting temperature test was conducted to investigate the influence of sulfur removal under the3.1.1. conditions Effect of employed,Roasting Temperature which were roasting time of 2 h and oxygen content of 80%, and the results are shown in Figure3. The roasting temperature test was conducted to investigate the influence of sulfur removal In the oxidizing roasting process, the ignition point of pyrite was generally about 400 ◦C, but the under the conditions employed, which were roasting time of 2 h and oxygen content of 80%, and the temperature fluctuated with the chemical composition of the material. When the temperature rose to results are shown in Figure 3. 900 ◦C, the sulfur content in the oxidizing roasting ore was the lowest (10.52%), and the desulfurization In the oxidizing roasting process, the ignition point of pyrite was generally about 400 °C, but the rate was 71.24%. When the temperature increased from 900 ◦C to 1000 ◦C, the sulfur content in temperature fluctuated with the chemical composition of the material. When the temperature rose to the oxidized roasted ore increased to 14.36%, and the desulfurization rate was 60.74%. Therefore, 900 °C, the sulfur content in the oxidizing roasting ore was the lowest (10.52%), and the the oxidizing roasting temperature of 900 ◦C was reasonable. desulfurization rate was 71.24%. When the temperature increased from 900 °C to 1000 °C, the sulfur content in the oxidized roasted ore increased to 14.36%, and the desulfurization rate was 60.74%. Therefore, the oxidizing roasting temperature of 900 °C was reasonable.

Processes 2020, 8, 200 7 of 18

100 34 Desulfuriation rate 32 Sulfur content 90 Processes 2020, 8, 200 30 7 of 18 Processes 2020, 8, 200 28 80 7 of 18 26 24 10070 3422 Desulfuriation rate 32 20 Sulfur content 9060 3018

Content (%) 2816 8050 2614 Desulfuriation rate (%) 2412 7040 2210 30 208 60 186

Content (%) 5020 16300 400 500 600 700 800 900 1000 1100 1200 14 Roasting temperature (℃) Desulfuriation rate (%) 12 40 10 Figure 3. Effect of oxidizing roasting temperature on sulfur30 removal. 8 6 20 3.1.2. Effect of Roasting Time300 400 500 600 700 800 900 1000 1100 1200 Roasting temperature (℃) Tests with different roasting times were carried out to investigate the influence of sulfur removal under the conditionsFigure employed, 3. EffectEffect which of oxidizing were roasting roasting temperaturetemperature temperature onon sulfur sulfurof 900 removal.removal. °C and oxygen content of 80%, and the results are shown in Figure 4. 3.1.2. Effect of Roasting Time 3.1.2.Prolonging Effect of Roasting the roasting Time time was conducive to the removal of sulfur, but when the roasting time was Testsmore withthan di3.0fferent h, the roasting sulfur content times were in the carried roasting out ore to investigate increased theregularly. influence Therefore, of sulfur 3.0 removal h was Tests with different roasting times were carried out to investigate the influence of sulfur removal undersuitable the for conditions oxidizing employed, roasting, and which 4.85% were sulfur-bearing roasting temperature oxidizing of roasting 900 C andore oxygencould be content obtained of under the conditions employed, which were roasting temperature of 900 ◦°C and oxygen content of 80%,with anda desulfurization the results are rate shown of 86.74%. in Figure 4. 80%, and the results are shown in Figure 4.

Prolonging the roasting24 time was conducive to the removal of sulfur,100 but when the roasting time Desulfuriation rate was more than 3.0 h, the sulfur22 content in the roasting ore increased regularly.95 Therefore, 3.0 h was Sulfur content suitable for oxidizing roasting,20 and 4.85% sulfur-bearing oxidizing roasting90 ore could be obtained with a desulfurization rate18 of 86.74%. 85 16 80 24 100 14 75 Desulfuriation rate 22 95 12 Sulfur content 70 20 90 10 65 Content (%) Content 18 85 8 60

16 80 (%) rate Desulfuriation 6 55 14 75 4 50 12 70 2 45 10 65 0 40 Content (%) Content 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 8 60 Roasting time (h) 6 55 (%) rate Desulfuriation Figure4 4.4. EEffectffect of oxidizing roastingroasting timetime onon sulfursulfur removal.removal.50 2 45 Prolonging the roasting time was conducive to the removal of sulfur, but when the roasting time 3.1.3. Effect of Roasting Atmosphere0 40 was more than 3.0 h, the sulfur0.0 content 0.5 1.0 in 1.5 the roasting 2.0 2.5 ore 3.0 increased 3.5 4.0 regularly. 4.5 Therefore, 3.0 h was Roasting time (h) suitableTests for under oxidizing various roasting, atmospheric and 4.85% conditions sulfur-bearing were oxidizing conducted roasting to investigate ore could bethe obtained influence with of asulfur desulfurization removal under rate of Figurethe 86.74%. conditions 4. Effect of employed, oxidizing roasting which time were on roastingsulfur removal. temperature of 900 °C and roasting time of 3.0 h, and the results are shown in Figure 5. 3.1.3.3.1.3.It EEffect ffisect known of Roasting from Figure AtmosphereAtmosphere 5 that the increase in oxygen content had an obvious effect on the removal of sulfur. With the increase of oxygen content, the sulfur content in the oxidizing roasting TestsTests underunder various various atmospheric atmospheric conditions conditions were were conducted conducted to investigate to investigate the influence the influence of sulfur of ore decreased gradually and tended to be stable at 70% oxygen or higher. This shows that the Co- removalsulfur removal under the under conditions the conditions employed, employed, which were wh roastingich were temperature roasting temperature of 900 ◦C and of roasting 900 °C timeand bearing sulfur concentrate needs an oxygen-enriched atmosphere for oxidizing roasting. The oxygen ofroasting 3.0 h, andtime the of results3.0 h, and are the shown results inFigure are shown5. in Figure 5. content was 80%, the sulfur content in the oxidizing roasting ore was 0.63%, and the desulfurization ItIt isis known known from from Figure Figure5 that 5 that the increasethe increase in oxygen in oxygen content content had an had obvious an obvious e ffect on effect the removal on the rate was 98.28%. Compared with the sulfur content in the Co-bearing sulfur concentrate, the ofremoval sulfur. of With sulfur. the increase With the of increase oxygen content,of oxygen the cont sulfurent, content the sulfur in the content oxidizing in the roasting oxidizing ore decreased roasting graduallyore decreased and gradually tended to and be stable tended at 70%to be oxygenstable at or 70% higher. oxygen This or shows higher. that This the shows Co-bearing that the sulfur Co- concentratebearing sulfur needs concentrate an oxygen-enriched needs an oxygen-enric atmospherehed for atmosphere oxidizing roasting.for oxidizing The roasting. oxygen content The oxygen was 80%,content the was sulfur 80%, content the sulfur in the content oxidizing in the roasting oxidizing ore roasting was 0.63%, ore was and 0.63%, the desulfurization and the desulfurization rate was rate was 98.28%. Compared with the sulfur content in the Co-bearing sulfur concentrate, the

Processes 2020, 8, 200 8 of 18

ProcessesProcesses 2020 2020, ,8 8, ,200 200 88 ofof 1818 98.28%. Compared with the sulfur content in the Co-bearing sulfur concentrate, the desulfurization desulfurizationedesulfurizationffect was remarkable. effect effect was was The remarkable. remarkable. sulfides eff ectivelyThe The sulfides sulfides transformed effectively effectively into transformed transformed oxides, and into into favorable oxides, oxides, conditions and and favorable favorable were conditionscreatedconditions for werewere the separation createdcreated forfor and thethe extraction separationseparation of andand cobalt extractionextraction and iron. ofof cobaltcobalt andand iron.iron.

2020 120120 Desulfuriation Desulfuriation rate rate 1818 Sulfur Sulfur content content 110110 1616 100100 1414

1212 9090

1010 8080 88 Content (%) Content Content (%) Content 66 7070

4 Desulfuriation rate (%) 4 Desulfuriation rate (%) 6060 22 5050 00

-2-2 4040 1010 20 20 30 30 40 40 50 50 60 60 70 70 80 80 90 90 100 100 OxygenOxygen content content (%) (%)

FigureFigure 5. 5. Effect EffectEﬀect of of oxygen oxygen content content on on sulfur sulfur removal. removal.

3.1.4. Analysis and Characterization of Oxidizing Roasting Ores 3.1.4.3.1.4. AnalysisAnalysis andand CharacterizationCharacterization ofof OxidizingOxidizing RoastingRoasting OresOres Repeated tests were conducted using the comprehensive conditions employed, which were RepeatedRepeated teststests werewere conductedconducted using using thethe comprecomprehensivehensive conditionsconditions employed,employed, which which werewere Co-Co- Co-bearing sulfur concentrate mass of 500 g, oxidizing roasting temperature of 900 C, oxidizing bearingbearing sulfur sulfur concentrate concentrate mass mass of of 500 500 g, g, oxidizing oxidizing roasting roasting temperature temperature of of 900 900 °C, °C, oxidizing oxidizing◦ roasting roasting roasting time of 3.0 h, and oxygen content of 80%. Moreover, the multiple oxidizing roasting ore times timetime ofof 3.03.0 h,h, andand oxygenoxygen contentcontent ofof 80%.80%. Moreover,Moreover, thethe multiplemultiple oxidizingoxidizing roastingroasting oreore timestimes werewere were merged into a comprehensive sample for analysis and characterization, and cobalt and iron were mergedmerged intointo aa comprehensivecomprehensive samplesample forfor analysisanalysis andand characterization,characterization, andand cobaltcobalt andand ironiron werewere extracted by magnetic separation of samples. The main chemical composition of oxidizing roasting extractedextracted byby magneticmagnetic separationseparation ofof samples.samples. TheThe mainmain chemicalchemical compositioncomposition ofof oxidizingoxidizing roastingroasting ores is shown in Table6, and the main minerals of oxidizing roasting ores by X-ray di ﬀraction (XRD) oresores is is shown shown in in Table Table 6, 6, and and the the main main minerals minerals of of oxidizing oxidizing roasting roasting ores ores by by X-ray X-ray diffraction diffraction (XRD) (XRD) are shown in Figure6. areare shownshown inin FigureFigure 6.6.

Table 6. Main chemical composition of oxidizing roasting ores (%). TableTable 6. 6. Main Main chemical chemical composition composition of of oxidizing oxidizing roasting roasting ores ores (%). (%).

Fe Co S CaO MgO Al2O3 SiO2 Mn K2O Na2O TiO2 FeFe CoCo SS CaOCaO MgOMgO AlAl22OO33 SiOSiO22 MnMn KK22OO NaNa22OO TiOTiO22 35.2735.2735.27 0.76 0.76 0.620.62 0.62 5.12 5.12 5.12 3.72 3.72 3.72 8.71 8.71 8.71 8.34 8.34 8.34 0.04 0.040.04 0.710.710.71 0.25 0.25 0.25 2.13 2.13 2.13

18001800 1-Fe1-FeOO 11 22 33 16001600 11 2-CoFe2-CoFe22OO44 3-Fe O 14001400 3-Fe33O44 4-Mg4-Mg33SiSi44OO1010(OH)(OH)22 12001200 5-Ca5-Ca33FeFe33(SiO(SiO44))33 10001000

800800 11 Intensity (a.u.) Intensity (a.u.) 600600 11 11 4 400400 4 22 22 2211 4 11 4 55 33 22 200200 55 44

-5-5 0 0 5 5 10 10 15 15 20 20 25 25 30 30 35 35 40 40 45 45 50 50 55 55 60 60 65 65 70 70 75 75 80 80 85 85 90 90 22θθ (°) (°) FigureFigure 6. 6. XRD XRDXRD diffractogram diffractogram diﬀractogram of of oxidizing oxidizing roasting roasting ores. ores.

ResultsResults inin TableTable 66 showshow thatthat thethe sulfursulfur contentcontent inin thethe oxidizingoxidizing roastingroasting oreore obtainedobtained fromfrom thethe repeatedrepeated testtest underunder comprehensivecomprehensive conditionsconditions waswas 0.62%,0.62%, whichwhich isis basicallybasically consistentconsistent withwith thethe desulfurizationdesulfurization effect effect of of the the single single-condition-condition test test and and also also shows shows that that the the obtained obtained oxidizing oxidizing roasting roasting

Processes 2020, 8, 200 9 of 18 conditions are repeatable. XRD mineral composition showed that, after the oxidizing roasting of cobalt-bearing sulfur concentrate, cobalt changed from Co-pyrite and linneite to a CoFe2O4- dominated new cobalt mineral phase, and iron changed from pyrite to Fe2O3 and an Fe3O4-dominated new iron mineral phase. At the same time, new gangue mineral phases such as Mg3Si4O10 (OH)2 and Ca3Fe3(SiO4)3 also appeared.

3.2. Cobalt Extraction from Oxidizing Roasting Ores by Segregation Roasting–Magnetic Separation Processes 2020, 8, 200 9 of 18 3.2.1. Effect of Chlorinating Agent Dosage Calcium chloride and sodium chloride tests with varies dosages were conducted to investigate Results in Table6 show that the sulfur content in the oxidizing roasting ore obtained from the the influences of cobalt grade and recovery. The conditions employed were roasting temperature of repeated test under comprehensive conditions was 0.62%, which is basically consistent with the 900 °C, roasting time of 60 min, coke dosage of 8%, coke size of <0.6 mm, magnetic separation desulfurization effect of the single-condition test and also shows that the obtained oxidizing roasting grinding fineness of <60 μm occupying 85%, and magnetic separation field intensity H1 = 0.06 T, and conditions are repeatable. XRD mineral composition showed that, after the oxidizing roasting of the results are shown in Figure 7. cobalt-bearing sulfur concentrate, cobalt changed from Co-pyrite and linneite to a CoFe2O4-dominated During the segregation roasting process, the chlorinator reacts with SiO2, Al2O3, and other acid new cobalt mineral phase, and iron changed from pyrite to Fe2O3 and an Fe3O4-dominated new oxides and H2O in the ore to generate hydrogen chloride gas. The valuable metal cobalt and iron react iron mineral phase. At the same time, new gangue mineral phases such as Mg3Si4O10 (OH)2 and with hydrogen chloride to produce CoCl2 and FeCl2. Because CoCl2 and FeCl2 are unstable, they are Ca3Fe3(SiO4)3 also appeared. easily reduced to metal particles by H2 and adsorb on the surface of coke; meanwhile, hydrogen chloride3.2. Cobalt gas Extraction will react from with Oxidizing its elements Roasting in chlorination, Ores by Segregation with too Roasting–Magneticmuch or too little dosage, Separation which will affect the separation and extraction of cobalt and iron. The grade and recovery of cobalt concentrate can3.2.1. be E improvedffect of Chlorinating by increasing Agent the Dosage amount of calcium chloride or sodium chloride [15]. When the amountCalcium of calcium chloride chloride and sodiumwas 6%, chloride the highest tests cobalt with grade varies was dosages 6.22%, were but conductedthe lowest torecovery investigate was 66.14%;the influences when the of cobalt amount grade of calcium and recovery. chloride The was conditions 12%, the employedhighest recovery were roasting of cobalt temperature was 82.11%, of and the cobalt grade was 4.12%. When sodium chloride was used as the chlorinating agent, the grade 900 ◦C, roasting time of 60 min, coke dosage of 8%, coke size of <0.6 mm, magnetic separation grinding of cobalt was lower than 3%, and the recovery of cobalt was lower with the increase of dosage. The fineness of <60 µm occupying 85%, and magnetic separation field intensity H1 = 0.06 T, and the results resultsare shown showed in Figure that 7calcium. chloride was suitable as chlorinating agent, taking into account cobalt grade and cobalt recovery, and the amount of calcium chloride was 12%.

7.0 100 7.0 100 (a)-CaCl 6.5 2 Cobalt grade 96 6.5 (b)-NaCl Cobalt grade 96 6.0 Cobalt recovery 92 6.0 Cobalt recovery 92 5.5 88 5.5 88 84 84 5.0 5.0 80 80 4.5 4.5 76 76 4.0 4.0 72 72 3.5 3.5 68 68 3.0 3.0

Grade (%) Grade 64 (%) Grade 64

2.5 (%) Recovery 2.5 (%) Recovery 60 60 2.0 2.0 56 56 1.5 52 1.5 52 1.0 48 1.0 48 0.5 44 0.5 44 0.0 40 0.0 40 5 6 7 8 9 10 11 12 13 14 15 5 6 7 8 9 10 11 12 13 14 15 Dosage (%) Dosage (%) FigureFigure 7. 7. EffectEﬀect of of c chlorinatinghlorinating a agentgent d dosageosage on on cobalt cobalt extraction. extraction.

3.2.2. DuringEffect of the Roasting segregation Temperature roasting process, the chlorinator reacts with SiO2, Al2O3, and other acid oxides and H2O in the ore to generate hydrogen chloride gas. The valuable metal cobalt and iron react Tests with different temperatures were conducted to investigate the influence of cobalt grade with hydrogen chloride to produce CoCl2 and FeCl2. Because CoCl2 and FeCl2 are unstable, they are and recovery. The conditions employed were roasting time of 60 min, calcium chloride dosage of easily reduced to metal particles by H2 and adsorb on the surface of coke; meanwhile, hydrogen 12%,chloride coke gas dosage will react of 8%, with coke its elementssize of < in0.6 chlorination, mm, magnetic with separation too much orgrinding too little fineness dosage, of which < 60 willμm occupyingaﬀect the separation 85%, and andmagnetic extraction separation of cobalt field and intensity iron. The H1 grade = 0.06 and T, recoveryand the results of cobalt are concentrate shown in Figurecan be 8. improved by increasing the amount of calcium chloride or sodium chloride [15]. When the amountFigure of calcium8 shows chloride that the was roasting 6%, the temperature highest cobalt was grade also was one 6.22%, of the but key the factors lowest to recovery determine was whether66.14%; whenthe separation the amount process of calcium was chloridegoing on. was Raisin 12%,g thethe highesttemperature recovery is conducive of cobalt was to the 82.11%, reaction and andthe cobaltaccelerating grade wasthe 4.12%.reaction When rate. sodiumIf the temperat chlorideure was is usedtoo low as the to chlorinatingreach the critical agent, point the grade of the of chemicalcobalt was reaction lower than of the 3%, ore, and this the will recovery result ofin cobaltslow or was difficult lower chemical with the increasereaction ofprocess dosage. [16,17]. The results With showed that calcium chloride was suitable as chlorinating agent, taking into account cobalt grade and cobalt recovery, and the amount of calcium chloride was 12%.

3.2.2. Effect of Roasting Temperature Tests with different temperatures were conducted to investigate the influence of cobalt grade and recovery. The conditions employed were roasting time of 60 min, calcium chloride dosage of 12%, coke dosage of 8%, coke size of < 0.6 mm, magnetic separation grinding fineness of < 60 µm occupying 85%, and magnetic separation field intensity H1 = 0.06 T, and the results are shown in Figure8. Processes 2020, 8, 200 10 of 18 the increase of roasting temperature, the grade of cobalt increased, and the recovery of cobalt increased first and then decreased. When the temperature increased to 950 °C, the grade of cobalt increased to 7.63%, and the recovery rate of cobalt was 85.47%; when the temperature increased to 1000 °C, the grade of cobalt increased to 8.23%, but the recovery rate of cobalt decreased to 71.81%. At the same time, it was found that when the temperature of segregation roasting was 1000 °C, the segregation roasting ore was in a melting state, and obvious metal particles appeared, so it was difficult to grind the material to a particle size less than 60 μm. Therefore, 950 °C was suitable for segregationProcesses 2020 ,roasting,8, 200 and cobalt concentrate with cobalt grade of 7.63% and recovery of 85.47% could10 of 18 be obtained.

10 100 Cobalt grade Processes 2020, 8, 200 9 96 10 of 18 Cobalt recovery 8 92 the increase of roasting temperature, the grade of cobalt increased, and the recovery of cobalt 7 88 increased first and then decreased. When the temperature increased to 950 °C, the grade of cobalt 6 84 increased to 7.63%, and the recovery rate of cobalt was 85.47%; when the temperature increased to 1000 °C, the grade of cobalt5 increased to 8.23%, but the recovery rate of 80cobalt decreased to 71.81%.

At the same time, it was foundGrade (%) 4 that when the temperature of segregation76 roasting was 1000 °C, the Recovery (%) segregation roasting ore was3 in a melting state, and obvious metal particles72 appeared, so it was difficult to grind the material2 to a particle size less than 60 μm. Therefore,68 950 °C was suitable for segregation roasting, and cobalt concentrate with cobalt grade of 7.63% and recovery of 85.47% could 1 64 be obtained. 0 60 750 800 850 900 950 1000 1050 10 Roasting temperature (℃) 100 Cobalt grade 9 96 Cobalt recovery FigureFigure 8. 8. EffectEﬀect of of segregation segregation roasting roasting te temperaturemperature on on cobalt cobalt extraction. extraction. 8 92

Figure8 shows that the roasting7 temperature was also one of the key factors88 to determine whether 3.2.3. Effect of Roasting Time the separation process was6 going on. Raising the temperature is conducive84 to the reaction and acceleratingTests with the different reaction segregation rate. If the temperatureroasting times is toowere low conducted to reach the to criticalinvestigate point the of influence the chemical of 5 80 cobaltreaction grade of theand ore, recovery. this will The result conditions in slow employed or diﬃcult were chemical roasting reaction temperature process of [95016,17 °C,]. Withcalcium the

Grade (%) 4 76 chlorideincrease dosage of roasting of 12%, temperature, coke dosage the gradeof 8%, of coke cobalt size increased, of <0.6 mm, and themagnetic recoveryRecovery (%) separation of cobalt grinding increased 3 72 finenessfirst and of then <60 μ decreased.m occupying When 85%, the and temperature magnetic separation increased field to 950 intensity◦C, the H grade1 = 0.06 of T, cobalt and the increased results areto 7.63%,shown andin Figure the recovery 9. 2 rate of cobalt was 85.47%; when the temperature68 increased to 1000 ◦C, the grade of cobalt increased1 to 8.23%, but the recovery rate of cobalt decreased64 to 71.81%. At the same 10 100 Cobalt grade time, it was found that when0 the temperature of segregation roasting was60 1000 ◦C, the segregation 9 96 roasting ore was in a melting750 state, 800 and obvious 850 metal 900 particles 950 Cobalt recovery 1000 appeared, 1050 so it was difficult to grind 8 Roasting temperature (℃) 92 the material to a particle size less than 60 µm. Therefore, 950 ◦C was suitable for segregation roasting, and cobalt concentrateFigure with8. Effect7 cobalt of segregation grade of 7.63% roasting and temperature recovery ofon 85.47%cobalt88 extraction. could be obtained. 6 84 3.2.3. Effect of Roasting Time 3.2.3. Effect of Roasting Time5 80

TestsTests with with different different Grade (%) segregation 4 roasting roasting times times were were conducted conducted to to76 investigate investigate the the influence influence of of Recovery (%) cobaltcobalt grade grade and and recovery. recovery. 3The The conditions conditions employed employed were were roasting roasting temperature temperature72 of of 950 950 °C,◦C, calcium calcium chloride dosage of 12%, coke dosage of 8%, coke size of <0.6 mm, magnetic separation grinding chloride dosage of 12%, coke2 dosage of 8%, coke size of <0.6 mm, magnetic68 separation grinding µ finenessfineness of of <60<60 μmm occupying occupying 85%, 85%, and and magnetic magnetic separation separation field field intensity H1 == 0.060.06 T, T, and and the the results results 1 64 areare shown shown in in Figure Figure 99.. 0 60 40 50 60 70 80 90 100 110 10 Roasting time (min) 100 Cobalt grade 9 96 Cobalt recovery Figure 9. Effect of segregation roasting time on cobalt extraction. 8 92

Results in Figure 9 show7 that the cobalt grade and recovery rate88 increased first and then decreased as time continued.6 The reason for this phenomenon is that the84 chemical reactions in the segregation roasting process5 are mainly divided into decomposition, chlorination,80 and reduction, but

Grade (%) 4 76 Recovery (%) 3 72

2 68

1 64

0 60 40 50 60 70 80 90 100 110 Roasting time (min)

FigureFigure 9. 9. EffectEﬀect of of segregation segregation roasting roasting time time on on cobalt cobalt extraction. extraction.

Results in Figure 9 show that the cobalt grade and recovery rate increased first and then decreased as time continued. The reason for this phenomenon is that the chemical reactions in the segregation roasting process are mainly divided into decomposition, chlorination, and reduction, but

Processes 2020, 8, 200 11 of 18

decomposition, chlorination, reduction, and other reactions are not only sequential chemical reactions, they are also complex phase changes. The segregation roasting time mainly affects the Processesdegree2020, of8, 200chemical reaction in the process of segregation roasting. The longer the segregation roasting11 of 18 time, the more the chemical reaction is complete. However, the shorter the segregation roasting time, the less complete the effective chemical reaction is [18,19]. For this, it is difficult to achieve the ideal Resultsseparation in Figureeffect. Considering9 show that that the the cobalt separation grade androasting recovery time was rate 75 increased min, a cobalt first concentrate and then decreased with as timea cobalt continued. grade of The 7.72% reason and a for cobalt this recovery phenomenon of 86.79% is thatcould the be chemicalobtained. reactions in the segregation roasting process are mainly divided into decomposition, chlorination, and reduction, but decomposition, chlorination,3.2.4. Effect reduction, of Coke Dosage and other reactions are not only sequential chemical reactions, they are also complex phaseTests with changes. varies Thecoke segregationdosages were roasting carried out time to mainlyinvestigate affects the influence the degree of cobalt of chemical grade and reaction in therecovery. process ofThe segregation conditions roasting.employed Thewere longer roasting the te segregationmperature of roasting 950 °C, time,roasting the time more of the75 min, chemical reactioncalcium is complete. chloride dosage However, of 12%, the shortercoke size the of segregation<0.6 mm, magnetic roasting separation time, the grinding less complete fineness the of e<60ffective chemicalμm occupying reaction is85%, [18 ,and19]. magnetic For this, separation it is difficult field to intensity achieve H the1 = 0.06 ideal T, and separation the results eff ect.are shown Considering in Figure 10. that the separation roasting time was 75 min, a cobalt concentrate with a cobalt grade of 7.72% and a Figure 10 reveals that the solid reducing agent was mainly used for coke, lignite, anthracite, cobalt recovery of 86.79% could be obtained. bituminous coal, and charcoal, among which coke was most commonly used. Because of more impure elements in coal, this will produce certain ash and affect the product quality. Charcoal from 3.2.4. Effect of Coke Dosage wood is generally used for beneficiation or smelting of precious metals. When the amount of coke Testsincreased with to varies 10%, the coke highest dosages cobalt were grade carried was 7.52% out to, and investigate the recovery the influencerate of cobalt ofcobalt was 87.85%. grade and When the amount of coke exceeded 10%, the grade of cobalt decreased obviously, and the recovery recovery. The conditions employed were roasting temperature of 950 ◦C, roasting time of 75 min, calciumrate chlorideof cobalt dosageincreased of to 12%, a certain coke extent. size of Beca<0.6use mm, coke magneticis a reduction separation carrier for grinding volatile finenessmetal of chloride in the process of chlorination segregation roasting, and it adjusts the atmosphere of <60 µm occupying 85%, and magnetic separation field intensity H = 0.06 T, and the results are shown chlorination and segregation properly, too much or too little coke 1will affect the grade and recovery in Figure 10. of cobalt. Therefore, it was reasonable to use 10% coke.

10.0 100 Cobalt grade 9.5 96 Cobalt recovery 9.0 92

8.5 88

8.0 84

7.5 80 76 7.0 72 6.5 Grade (%)

68 Recovery (%) 6.0 64 5.5 60 5.0 56 4.5 52 4.0 4 5 6 7 8 9 10 11 12 13 14 15 16 Dosgae (%)

FigureFigure 10. 10.E ﬀEffectect of of cokecoke dosage on on cobalt cobalt extraction. extraction.

Figure3.2.5. Effect 10 reveals of Coke thatSize the solid reducing agent was mainly used for coke, lignite, anthracite, bituminousA coke coal, size and test charcoal, was conducted among which to investigate coke was its mostinfluence commonly on cobalt used. grade Because and recovery. of more The impure elementsconditions in coal, employed this will were produce oxidizing certain roasting ash and ore a ffmassect theof 50 product g, roasting quality. temperature Charcoal of from 950 wood°C, is generallyroasting used time for of beneficiation 75 min, calcium or smeltingchloride dosage of precious of 12%, metals. coke dosage When of the 10%, amount magnetic of cokeseparation increased to 10%,grinding the highest fineness cobalt of <60 gradeμm occupying was 7.52%, 85%, andand magnetic the recovery separation rate offield cobalt intensity was H 87.85%.1 = 0.06 T, When and the amountthe results of coke are exceeded shown in Figure 10%,the 11. grade of cobalt decreased obviously, and the recovery rate of cobalt increasedThe specific to a surface certain area extent. (as = As/m) Because is an coke important is a reduction index in measuring carrier for the volatile adsorption metal capacity chloride in the processof coke. ofThe chlorination smaller the coke segregation particle size, roasting, the stro andnger itits adjusts adsorption the capacity. atmosphere Figure of 11 chlorination shows that and with the decrease of coke particle size, the grade of cobalt increased, and the recovery of cobalt segregation properly, too much or too little coke will affect the grade and recovery of cobalt. Therefore, increased first and then decreased. In the process of segregation roasting, the reductant can adjust it was reasonable to use 10% coke. the atmosphere for chloride separation and absorb and reduce the volatile metal chloride [19,20]. 3.2.5. Effect of Coke Size A coke size test was conducted to investigate its influence on cobalt grade and recovery. The conditions employed were oxidizing roasting ore mass of 50 g, roasting temperature of 950 ◦C, roasting time of 75 min, calcium chloride dosage of 12%, coke dosage of 10%, magnetic separation grinding fineness of <60 µm occupying 85%, and magnetic separation field intensity H1 = 0.06 T, and the results are shown in Figure 11. Processes 2020, 8, 200 12 of 18

Therefore, coke was used as the reducing agent and the carrier of adsorbing metal particles in the process of separation and roasting. Too much or too little coke is not conducive to the separation and Processes 2020extraction, 8, 200 of cobalt. The size of coke was < 0.4 mm, which was suitable to obtain a cobalt concentrate 12 of 18 with a cobalt grade of 9.03% and a cobalt recovery of 87.77%.

12.0 100 Cobalt grade 11.5 95 11.0 Cobalt recovery 90 10.5 85 10.0 9.5 80

9.0 75 8.5 70 8.0 65 7.5

Grade (%) 60 Processes 2020, 8, 200 7.0 Recovery (%) 12 of 18 6.5 55 6.0 50 Therefore, coke was used as5.5 the reducing agent and the carrier of adsorbing metal particles in the 45 process of separation and roasting.5.0 Too much or too little coke is not conducive to the separation and 4.5 40 extraction of cobalt. The size4.0 of coke was < 0.4 mm, which was suitable to35 obtain a cobalt concentrate with a cobalt grade of 9.03% -1.2and -1.1 a cobalt -1.0 -0.9 recovery -0.8 -0.7 -0.6 of 87.77%. -0.5 -0.4 -0.3 -0.2 -0.1 0.0 Size (mm)

12.0 100 Cobalt grade Figure11.5 Figure 11. 11.Eff Effectect of of coke coke sizesize on on cobalt cobalt extraction. extraction.95 11.0 Cobalt recovery 90 10.5 3.2.6. Effect of Magnetic Separation Conditions of Segregation Roasting 85Ores The specific surface area (a10.0s = As/m) is an important index in measuring the adsorption capacity of coke. The smallerThe grinding the coke fineness particle9.5 and size,magnetic the strongerfield intensity its adsorptionare important capacity. factors80 that Figure affect 11 cobalt shows grade that with 9.0 75 the decreaseand cobalt of coke recovery particle in size,the8.5 magnetic the grade separation of cobalt process increased, of segregation and the recovery roasting ofores. cobalt Reasonable increased first 70 8.0 and thengrinding decreased. fineness In theand processmagnetic offiel segregationd intensity are roasting, beneficial the to improve reductant the65 canseparation adjust and the extraction atmosphere for 7.5

index of cobalt. MagneticGrade (%) separation fineness grinding of segregation roasting60 ores was conducted to chloride separation and absorb7.0 and reduce the volatile metal chloride [Recovery (%) 19,20]. Therefore, coke was investigate its influence on6.5 cobalt grade and recovery, and the conditions55 employed were segregation used as the reducing agent and6.0 the carrier of adsorbing metal particles in the process of separation roasting temperature of 950 °C, segregation roasting time of 75 min, calcium50 chloride dosage of 12%, 5.5 and roasting. Too much or too little coke is not conducive to the separation45 and extraction of cobalt. coke dosage of 10%, coke size5.0 of <0.4 mm, and magnetic separation field intensity H1 = 0.06 T, and the The sizeresults of coke are was shown< 0.4 in Figure mm,4.5 which 12a. was suitable to obtain a cobalt concentrate40 with a cobalt grade of 4.0 35 20 -1.2 -1.1 -1.0 -0.9 -0.8100 -0.7 -0.6 -0.520 -0.4 -0.3 -0.2 -0.1 0.0 100 9.03% and a cobalt(a) recovery of 87.77%. Cobalt grade 98 (b) Cobalt grade 98 18 Size (mm) 18 Cobalt recovery 96 Cobalt recovery 96 94 94 3.2.6. Effect of16 Magnetic Separation Conditions of Segregation16 Roasting Ores Figure 11. Effect of 92coke size on cobalt extraction. 92 14 90 14 90 The grinding fineness and magnetic field88 intensity are important factors that affect88 cobalt grade 12 12 3.2.6. Effect of Magnetic Separation Conditions86 of Segregation Roasting Ores 86 and cobalt recovery in the magnetic separation84 process of segregation roasting ores.84 Reasonable 10 10

The(%) Grade grinding fineness and magnetic field82 intensityGrade (%) are important factors that affect cobalt 82grade Recovery (%) Recovery grinding fineness and magnetic field intensity are beneficial to improve the separation andRecovery (%) extraction and cobalt8 recovery in the magnetic separation80 process8 of segregation roasting ores. Reasonable80 78 78 index of cobalt.6 Magnetic separation fineness grinding of6 segregation roasting ores was conducted to grinding fineness and magnetic field intensity76 are beneficial to improve the separation and extraction76 investigate its influence on cobalt grade and recovery,74 and the conditions employed were74 segregation index of4 cobalt. Magnetic separation fineness grinding of4 segregation roasting ores was conducted to 72 72 roastinginvestigate temperature2 its influence of 950 on◦C, cobalt segregation grade androasting recove70 ry, timeand2 the of conditions 75 min, calcium employed chloride were segregation dosage70 of 12%, 70 75 80 85 90 95 100 0.02 0.04 0.06 0.08 0.10 0.12 0.14 coke dosageroasting of temperature 10%,Grinding coke finess sizeof 950 (＜ of 60 °C, μm< 0.4occupying segregation mm, ) and roasting magnetic time separationof 75 min,Manetic calcium field intensity intensity chloride field H (T) dosageH = 0.06of 12%, T, and the 1 1 results arecoke shown dosage of in 10%, Figure coke 12 sizea. of <0.4 mm, and magnetic separation field intensity H1 = 0.06 T, and the resultsFigure are shown 12. Effect in Figureof grinding 12a. fineness and magnetic intensity field on cobalt extraction: (a)-Grinding fineness; (b)-Magnetic intensity filed. 20 100 20 100 (a) Cobalt grade 98 (b) Cobalt grade 98 18 18 Grinding fineness is the key Cobalt factor recovery that96 reflects the degree of dissociation Cobalt of recovery the mineral96 94 94 16 16 monomer, and basic dissociation of the mineral92 is the primary factor to realize mineral separation.92 In the process14 of magnetic separation, it is beneficial90 to improve14 the cobalt grade of cobalt concentrate.90 88 88 12 12 Figure 12a shows that 90% of particles finer than86 60 μm were suitable for cobalt magnetic separation,86 84 84 and cobalt10 concentrate with a cobalt grade of 15.63% and10 cobalt recovery of 82.38% were obtained. If

Grade (%) Grade 82 Grade (%) 82 Recovery (%) Recovery Recovery (%) the grinding8 fineness is too high, the recovery80 rate of cobalt8 concentrate decreases significantly,80 and 78 78 the particle6 size becomes smaller. 6 76 76 74 74 4 4 72 72 2 70 2 70 70 75 80 85 90 95 100 0.02 0.04 0.06 0.08 0.10 0.12 0.14 Grinding finess (＜ 60 μm occupying ) Manetic intensity field H (T) 1 Figure 12.FigureEff ect12. Effect of grinding of grinding fineness fineness and and magnetic magnetic intensity intensity field field on cobalt on cobalt extraction: extraction: (a)-Grinding (a) Grinding fineness; (b) Magnetic intensity filed.fineness; (b)-Magnetic intensity filed.

GrindingGrinding fineness fineness is the is key the factor key factor that reflectsthat reflects the degreethe degree of dissociation of dissociation of theof the mineral mineral monomer, monomer, and basic dissociation of the mineral is the primary factor to realize mineral separation. In and basic dissociation of the mineral is the primary factor to realize mineral separation. In the process the process of magnetic separation, it is beneficial to improve the cobalt grade of cobalt concentrate. of magneticFigure separation, 12a shows that it 90% is beneficial of particles to finer improve than 60 theμm cobaltwere suitable grade for of cobalt cobalt magnetic concentrate. separation, Figure 12a shows thatand cobalt 90% ofconcentrate particles with finer a cobalt than 60gradeµm of were 15.63% suitable and cobalt for recovery cobalt magnetic of 82.38% were separation, obtained. and If cobalt the grinding fineness is too high, the recovery rate of cobalt concentrate decreases significantly, and the particle size becomes smaller.

Processes 2020, 8, 200 13 of 18 concentrate with a cobalt grade of 15.63% and cobalt recovery of 82.38% were obtained. If the grinding finenessProcesses is too 2020 high,, 8, 200 the recovery rate of cobalt concentrate decreases significantly, and the13 particleof 18 size becomes smaller.In the process of magnetic separation, the effect of external forces becomes more obvious. To Inimprove the process the magnetic of magnetic separation separation, index, thethe field e ffintensityect of externalof magnetic forces separation becomes needs more to be obvious. To improvechanged.the A magnetic magnetic separation separation field index, intensity the test field for the intensity segregation of magnetic roasting ores separation was conducted needs to be changed.to investigate A magnetic the separationinfluence of cobalt field intensitygrade and testrecovery. for the The segregation conditions employed roasting were ores oxidizing was conducted to investigateroasting theore mass influence of 50 g, of segregation cobalt grade roasting and temp recovery.erature of The 950 conditions°C, segregation employed roasting time were of 75 oxidizing min, calcium chloride dosage of 12%, coke dosage of 10%, coke size of <0.4 mm, and magnetic roastingseparation ore mass grinding of 50 g,fineness segregation of <60 μm roasting occupying temperature 90%, and the ofresults 950 are◦C, shown segregation in Figure roasting 12b. time of 75 min, calciumWith the chloride increase dosageof magnetic of 12%,field intensity, coke dosage the grade of 10%, of cobalt coke decreased size of and<0.4 the mm, recovery and of magnetic separationcobalt grinding increased. fineness In the process of <60 ofµ mmagnetic occupying separation, 90%, anddifferent the resultsmagnetic are minerals shown received in Figure the 12 b. Withcombined the increase action of of magnetic magnetic field field force, intensity, gravity, medium the grade resistance, of cobalt and decreasedother external and mechanical the recovery of cobalt increased.forces in the magnetic In the process field. When of magneticthe magnetic separation, field force of ditheff erentmineral magnetic particles is minerals greater than received the the mechanical external force, they will enter the magnetic products and become the concentrate; combined action of magnetic field force, gravity, medium resistance, and other external mechanical otherwise, they will enter the non-magnetic products and become the tailings. Therefore, the forces in the magnetic field. When the magnetic field force of the mineral particles is greater than magnetic field intensity H1 = 0.08 T was suitable, and the cobalt concentrate with cobalt grade of the mechanical15.13% and external cobalt recovery force, of they 90.04% will was enter obtained. the magnetic products and become the concentrate; otherwise, they will enter the non-magnetic products and become the tailings. Therefore, the magnetic 3.3. Iron Extraction from Cobalt Magnetic Separation Tailings by Magnetic Separation field intensity H1 = 0.08 T was suitable, and the cobalt concentrate with cobalt grade of 15.13% and cobalt recoveryThe magnetism of 90.04% of was new obtained. cobalt minerals and iron minerals was different in segregation roasting ores, and the effective extraction index of cobalt was obtained from segregation roasting ores by one- 3.3. Ironstage Extraction magnetic from separation. Cobalt Extraction Magnetic of Separation iron from cobalt Tailings magnetic by Magnetic separation Separation tailings was conducted by changing the magnetic field intensity using second-stage magnetic separation. Different magnetic Theseparation magnetism field ofintensity new cobalt (H2) tests minerals were investigated, and iron minerals and the test was results different are shown in segregation in Figure 13. roasting ores, and the effectiveThe iron extraction recovery rate index of iron of cobalt concentrate was obtainedimproved by from increasing segregation the magnetic roasting field ores intensity, by one-stage magneticand separation. the iron grade Extraction in Figure 13 of re ironveals froma regular cobalt increase magnetic as the magnetic separation field tailings intensity was decreased. conducted by changingThe the existing magnetic iron and field steel intensityenterprises using require second-stage that the iron grade magnetic of iron and separation. steel concentrate Different is more magnetic than 55%, so the magnetic field intensity is H2 = 0.14 T, which is suitable to obtain iron concentrate separation field intensity (H ) tests were investigated, and the test results are shown in Figure 13. with iron grade of 60.02%2 and iron recovery of 80.01%.

70 100 Iron grade 68 Iron recovery 95

66 90

64 85

62 80

60 75

Grade (%) Grade 58 70 Recovery (%)

56 65

54 60

52 55

50 50 0.08 0.10 0.12 0.14 0.16 0.18 0.20 Manetic intensity field H (T) 2 FigureFigure 13. 13.Eﬀ Effectect of of magnetic magnetic intensity intensity field ﬁeld on oniron iron extraction. extraction.

The3.4. iron The Entire recovery Co, Fe rate Extraction of iron Process concentrate Test improved by increasing the magnetic field intensity, and the ironThe grade entire in process Figure of 13cobalt reveals and iron a regular extraction increase was conducted as the magnetic to investigate field the intensity influence decreased.of The existingcobalt irongrade and and steelrecovery, enterprises and the conditions require that employed the iron were grade oxidizing of iron roasting and steeltemperature concentrate of 900 is more °C, oxidizing roasting time of 3.0 h, oxygen content of 80%, segregation roasting temperature of 950 than 55%, so the magnetic field intensity is H2 = 0.14 T, which is suitable to obtain iron concentrate °C, segregation roasting time of 75 min, calcium chloride dosage of 12%, coke dosage of 10%, coke with iron grade of 60.02% and iron recovery of 80.01%. size of <0.4 mm, magnetic separation grinding fineness of <60 μm occupying 90%, cobalt magnetic separation field intensity H1 = 0.08 T, and iron magnetic separation field intensity H2 = 0.14 T. The 3.4. The Entire Co, Fe Extraction Process Test results are shown in Tables 7–9. The entire process of cobalt and iron extraction was conducted to investigate the influence of cobalt grade and recovery, and the conditions employed were oxidizing roasting temperature of 900 ◦C, oxidizing roasting time of 3.0 h, oxygen content of 80%, segregation roasting temperature of 950 ◦C, Processes 2020, 8, 200 14 of 18 segregation roasting time of 75 min, calcium chloride dosage of 12%, coke dosage of 10%, coke size of <0.4 mm, magnetic separation grinding fineness of <60 µm occupying 90%, cobalt magnetic separation field intensity H1 = 0.08 T, and iron magnetic separation field intensity H2 = 0.14 T. The results are shown in Tables7–9.

Table 7. Results of oxidizing roasting–segregation roasting–magnetic separation (%).

Grade Recovery Products Productivity Co Fe Co Fe Cobalt concentrate 4.08 15.15 71.22 90.81 8.74 Iron concentrate 42.21 0.11 60.06 6.82 76.23 Tailings 53.71 0.03 9.31 2.37 15.03 Totals 100.00 0.68 33.26 100.00 100.00

Table 8. Main chemical composition of cobalt concentrate (%).

Co Fe S CaO MgO Al2O3 SiO2 Mn 15.15 71.22 0.04 1.22 0.36 0.55 3.12 0.002

Table 9. Main chemical composition of iron concentrate (%).

Fe Co S CaO MgO Al2O3 SiO2 Mn 60.06 0.11 0.05 8.03 4.11 0.56 6.56 0.006

Results in Tables7–9 show the extraction e ﬀects of cobalt and iron from Co-bearing sulfur concentrate in the Panxi area using the oxidizing roasting–segregation roasting–magnetic separation process. Cobalt concentrate with Co grade of 15.15%, Fe content of 71.22%, and cobalt recovery of 90.81% as well as iron concentrate with Fe grade of 60.06%, Co content of 0.11%, and iron recovery of 76.23% were obtained. The content of harmful elemental sulfur in the iron concentrate was also lower, which met the product quality requirements for highly eﬃcient extraction and separation of valuable metal cobalt and iron. Sulfur was produced in the form of SO2 in the oxidizing roasting process, which can be used as an important raw material for the preparation of sulfuric acid.

3.5. Cobalt and Iron Phase Transformation in the Segregation Roasting Process Analysis and characterization of segregation roasting ores, cobalt concentrate, and iron concentrate were investigated by X-ray diffraction (XRD), scanning electron microscopy (SEM), and energy spectroscopy (EDS) to find out the mineral phase evolution of cobalt minerals and iron minerals. XRD phase analysis results of cobalt concentrate and iron concentrate are shown in Figure 14. SEM and EDS images of cobalt and iron minerals are shown in Figure 15. According to the existing kinetic mechanism, the separation roasting stage can be divided into five stages. (1) Solid phase sintering reaction stage: oxidizing roasting, chlorinating agent, and reducing agent are dried at low temperature and then put into the baking furnace. Before chlorination segregation roasting, the pellet sintering process is shown, which is actually a series of solid-phase sintering reactions. (2) Decomposition stage: this stage is mainly the decomposition reaction of calcium chloride to form hydrogen chloride gas with strong activity, The reaction of the reducing agent and water vapor produces hydrogen and carbon monoxide. A catalyst promotes the positive reaction direction of the separation process to proceed rapidly. (3) Gas–solid reaction stage: the ore is still in the solid state because the separation roasting temperature is at 950 ◦C, which is not high enough for the softening point is more of a gas–solid reaction. Hydrogen chloride gas with strong volatilization activity reacts rapidly with metal in ores to form volatile chloride, and the decomposition pressure of various metal chlorides is different. The promotion and inhibition of catalysts and promoters can improve the volatilization rate of the metal cobalt target. (4) Reduction adsorption stage: the formation of volatile Processes 2020, 8, 200 15 of 18 metal chlorides through the gas–solid phase reaction forms a strong physical and chemical adsorption on the surface of the reducing agent. The decomposition pressures of various volatile metal chlorides are high and low, and the corresponding double effects of adsorption and reduction on the surface of Processesthe reducing 2020, 8, agent200 form fine metal particles. (5) New phase formation stage: the size of metal particles15 of 18 formed in the early stage of new phase formation is very fine, the corresponding specific surface area is the early stage makes the metal particles formed in the later stage continuously rearrange and pile large, and the fine particles themselves have strong adsorption, which belongs to the thermodynamic up, and the crystal particles grow, finally forming a certain size of metal particle compounds [21–25]. instability system. Then, the “metal core” in the early stage makes the metal particles formed in the After the separation roasting of oxidized roasted ore, cobalt changes from CoFe2O4 to a new later stage continuously rearrange and pile up, and the crystal particles grow, finally forming a certain cobalt mineral phase dominated by a [Co] Fe solid solution; iron changes from Fe2O3 to a new iron size of metal particle compounds [21–25]. mineral phase dominated by metal Fe and Fe3O4. Cobalt and iron belong to ferromagnetic metals. After the separation roasting of oxidized roasted ore, cobalt changes from CoFe O to a new cobalt Magnetite belongs to strong magnetic minerals, and their specific magnetization2 4coefficients are mineral phase dominated by a [Co] Fe solid solution; iron changes from Fe O to a new iron mineral different to some extent. The main minerals in the cobalt concentrate obtained2 3 from the separation phase dominated by metal Fe and Fe3O4. Cobalt and iron belong to ferromagnetic metals. Magnetite roasting ore after grinding were Fe, [Co]Fe, Fe3O4, and SiO2; the main minerals in the iron concentrate belongs to strong magnetic minerals, and their specific magnetization coefficients are different to some obtained from the second-stage magnetic separation were Fe3O4, FeO, Ca2Si2O4, and Ca2Al2O4. In the extent. The main minerals in the cobalt concentrate obtained from the separation roasting ore after process of magnetic separation, a small amount of gangue minerals such as SiO2, Ca2Si2O4, and grinding were Fe, [Co]Fe, Fe3O4, and SiO2; the main minerals in the iron concentrate obtained from Ca2Al2O4 entered into cobalt concentrate and iron concentrate as a result of mechanical entrainment. the second-stage magnetic separation were Fe O , FeO, Ca Si O , and Ca Al O . In the process of Therefore, there are many complex phase changes3 4 in the segregation2 2 4 roasting2 2 process.4 Combined magnetic separation, a small amount of gangue minerals such as SiO , Ca Si O , and Ca Al O entered with the analysis results, the possible chemical reactions in the 2segregation2 2 4 roasting2 process2 4 are into cobalt concentrate and iron concentrate as a result of mechanical entrainment. Therefore, there are shown in Equations (3)–(10). many complex phase changes in the segregation roasting process. Combined with the analysis results, the possible chemical reactionsCaCl in the2 + segregation xSiO2 + H2O roasting→ CaO·xSiO process2 + HCl are shown in Equations (3)–(10).(3)

CaClCaCl2 + Al2O+3·2SiOxSiO2·2H+ H2O O→ (CaO)CaO2·AlxSiO2O3·2SiO+ HCl2 + HCl (4) (3) 2 2 2 → · 2

CaCl + Al O 2SiOC 2H+ H2O → CO(CaO) + H2 Al O 2SiO + HCl(5) (4) 2 2 3· 2· 2 → 2· 2 3· 2 C + H2O CO + H2 (5) 3Fe2O3 + CO → 2Fe3O4 + CO2 (6) 3Fe O + CO 2Fe O + CO (6) 2 3 → 3 4 2 CoFe2O4 + 6HCl → CoCl2 + 2FeCl3 + 3H2O (7) CoFe O + 6HCl CoCl + 2FeCl + 3H O (7) 2 4 → 2 3 2 CoClCoCl2+ + HH2 → CoCo + +HClHCl (8) (8) 2 2 → FeClFeCl33+ + 3H22 → 2Fe ++ 6HCl6HCl (9) (9) → 2Co + 2FeCl3 + 3H2 2[Co]Fe + 6HCl (10) 2Co + 2FeCl3 + 3H2→→ 2[Co]Fe + 6HCl (10) In the system, the relationship between the diﬃculties in the reduction of iron cobalt metal chloride In the system, the relationship between the difficulties in the reduction of iron cobalt metal is FeCl3 < CoCl2, and cobalt is a kind of iron-soluble oxide. FeCl3 itself is a kind of chlorinating agent, chloride is FeCl3 < CoCl2, and cobalt is a kind of iron-soluble oxide. FeCl3 itself is a kind of chlorinating promoting the chlorination and segregation of cobalt, and forms the solid solution of [Co] Fe. agent, promoting the chlorination and segregation of cobalt, and forms the solid solution of [Co] Fe.

3000 2400 1-Fe Cobalt concentrate Iron concentrate 1-Fe3O4 2200 2-[Co]Fe 2700 3-Fe O 2-FeO 2000 1 3 4 2400 1 3-Ca2Si2O4 1800 4-SiO2 2100 4-Ca2Al2O4 1600 1 1800 1400 1200 1500 1000 1200 2

Intensity(a.u.) 800 (a.u.) Intensity 900 600 600 400 2 3 4 4 3 200 300

0 0 -5 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 -5 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 2θ (°) 2θ (°) FigureFigure 14. 14. XRDXRD diffractogram diﬀractogram of of the the cobalt cobalt co concentratencentrate and and iron iron concentrate. concentrate.

Processes 2020, 8, 200 16 of 18 Processes 2020, 8, 200 16 of 18

1200 1100 (a) 1000 O 900 Fe Mg 800 Fe Co 700 Ca

600 Al Co 500 400 Co Intensity (cps) Intensity 300 200 100 0 -100 012345 Binding energy (kev)

1200 (b) 1100 1000 Ca Fe O Si 900 800 700 600 Al 500 400

(cps) Intensity 300 Mg 200 100 0 -100 012345 Binding energy (kev)

2400 (c) Fe 2000

1600

1200

800 Intensity (cps) Intensity

Co 400 Co

-202468101214 Binding energy (kev) 300

(d) 250

200 Fe

150

100 (cps) Intensity O Fe O 50

-50 -0.50.00.51.01.52.02.53.03.54.04.55.0 Binding energy (kev)

FigureFigure 15. SEMSEM and and EDS EDS images images (a—segregation (a—segregation roasting roasting ores; ores; b—tailings;b—tailings; c—cobaltc—cobalt concentrate; concentrate; d— diron—iron concentrate). concentrate).

4.4. Conclusions (1)(1) TheThe refractoryrefractory Co-bearingCo-bearing sulfursulfur concentrateconcentrate sample was collected from a V-TiV-Ti magnetitemagnetite dressingdressing plantplant byby ﬂotationflotation inin thethe PanzhihuaPanzhihua areaarea ofof China,China, andand thethe Co-bearingCo-bearing sulfursulfur concentrateconcentrate containedcontained Co 0.68%, 0.68%, Fe Fe 33.26%, 33.26%, and and S 36.58%. S 36.58%. Pyrite Pyrite was the was main the sulfide main sulﬁde ore; Co-pyrite ore; Co-pyrite and linneite and linneitewere Co-bearing were Co-bearing minerals; minerals;gangue minerals gangue were minerals mica, werechlorite, mica, feldspar, chlorite, calcite, feldspar, etc., in calcite, Co-bearing etc., inconcentrate. Co-bearing concentrate. (2) A new process of oxidizing roasting–segregation roasting–magnetic separation is used to process the complex Co-bearing sulfur concentrate. The cobalt concentrate with Co of 15.15%, Fe 71.22%, and cobalt recovery of 90.81% as well as iron concentrate with Fe of 60.06%, Co of 0.11%, and

Processes 2020, 8, 200 17 of 18

(2) A new process of oxidizing roasting–segregation roasting–magnetic separation is used to process the complex Co-bearing sulfur concentrate. The cobalt concentrate with Co of 15.15%, Fe 71.22%, and cobalt recovery of 90.81% as well as iron concentrate with Fe of 60.06%, Co of 0.11%, and iron recovery of 76.23% were obtained under the comprehensive conditions of oxidizing roasting temperature of 900 °C, oxidizing roasting time of 3.0 h, oxygen content of 80%, segregation roasting time of 950 ◦C, segregation roasting time of 75 min, calcium chloride dosage of 12%, coke dosage of 10%, coke size of <0.4 mm, magnetic separation grinding ﬁneness of <60 µm occupying 90%, cobalt magnetic separation magnetic ﬁeld intensity of H1 = 0.08 T, and iron magnetic separation intensity of H2 = 0.14 T. Extraction of cobalt and iron is obvious. (3) The results of cobalt and iron phase transformations show that cobalt is transformed from Co-pyrite and linneite to a Co2FeO4-dominated new cobalt mineral phase. Iron is transformed from pyrite to Fe2O3 and an Fe3O4-dominated new iron mineral phase after oxidizing roasting. Sulfur is produced in the form of SO2 after oxidizing roasting; cobalt changed from CoFe2O4 to a new cobalt mineral phase dominated by [Co]Fe solid solution. Iron changed from Fe2O3 to a new iron mineral phase dominated by metal Fe and Fe3O4 after segregation roasting. (4) SEM, EDS, and XRD analyses of segregation roasting ores, cobalt concentrate, and iron concentrate show that the main minerals in cobalt concentrate are Fe, [Co] Fe, and a small amount of Fe3O4; the main minerals in iron concentrate are Fe3O4 and FeO; and there was a small amount of gangue minerals such as SiO2, Ca2Si2O4, and Ca2Al2O4 in the cobalt concentrate and iron concentrate because of mechanical entrainment.

Author Contributions: This is a joint work of the two authors; each author was in charge of their expertise and capability: J.X. for writing, formal analysis, and original draft preparation; Y.Z. for conceptualization, validation, methodology, and investigation. All authors have read and agreed to the published version of the manuscript. Funding: This work was supported by the Sichuan Science and Technology Program (Grant Nos.2019FS0451 and Nos.2018FZ0092); Key Laboratory of Sichuan Province for Comprehensive Utilization of Vanadium and Titanium Resources Foundation (2018FTSZ35); China Geological Big Survey (Grant No. DD20190694). Conﬂicts of Interest: The authors declare no conﬂict of interest. The funders had no role in the design, analyses, and interpretation of any data of the study.

References

1. Gulley, A.L.; Nassar, N.T.; Xun, S. China, the United States, and competition for resources that enable emerging technologies. Proc. Natl. Acad. Sci. USA 2018, 115, 4111–4115. [CrossRef][PubMed] 2. Yang, H.P.; Wang, W. Global Cobalt Resources Status and Exploitation Trends. Conserv. Util. Miner. Resour. 2019, 39, 41–49. (In Chinese) 3. Aleksandrov, P.V.; Medvedev, A.S.; Kadirov, A.A.; Imideev, V.A. Processing molybdenum concentrates using low-temperature oxidizing-chlorinating roasting. Russ. J. Non Ferr. Met. 2014, 55, 114–119. [CrossRef] 4. Aracena, A.; Jerez, O.; Ortiz, R.; Morales, J. Pyrite oxidation kinetics in an oxygen–nitrogen atmosphere at

temperatures from 400 to 500 ◦C. Can. Metall. Q. 2016, 55, 195–201. [CrossRef] 5. Aydin, F.; Ziyadanogullari, R. Separation of copper, cobalt and iron in solution obtained from copper ore. Asian J. Chem. 2007, 19, 5537. 6. Biswas, S.; Dey, R.; Mukherjee, S.; Banerjee, P.C. Bioleaching of nickel and cobalt from lateritic chromite overburden using the culture ﬁltrate of Aspergillus niger. Appl. Biochem. Biotechnol. 2013, 170, 1547–1559. [CrossRef] 7. Cui, F.; Mu, W.; Wang, S.; Xin, H.; Xu, Q.; Zhai, Y.; Luo, S. Sodium sulfate activation mechanism on co-sulfating roasting to nickel-copper sulﬁde concentrate in metal extractions, microtopography and kinetics. Miner. Eng. 2018, 123, 104–116. [CrossRef] 8. Cui, F.; Mu, W.; Wang, S.; Xin, H.; Shen, H.; Xu, Q.; Luo, S. Synchronous extractions of nickel, copper, and cobalt by selective chlorinating roasting and water leaching to low-grade nickel-copper matte. Sep. Purif. Technol. 2018, 195, 149–162. [CrossRef] 9. Erust, C.; Akcil, A. Copper and cobalt recovery from pyrite ashes of a sulphuric acid plant. Waste Manag. Res. 2016, 34, 527–533. [CrossRef] Processes 2020, 8, 200 18 of 18

10. Xiao, J.H.; Zhang, Y.S. Recovering Cobalt and Sulfur in Low Grade Co-bearing V–Ti Magnetite Tailings Using Flotation Process. Processes 2019, 7, 536. [CrossRef] 11. Morcali, M.H.; Khajavi, L.T.; Dreisinger, D.B. Extraction of nickel and cobalt from nickeliferous limonitic laterite ore using borax containing slags. Int. J. Miner. Process. 2017, 167, 27–34. [CrossRef]

12. Wang, W.; Zhang, Y.; Liu, X.; Xu, S. A simplified process for recovery of Li and Co from spent LiCoO2 cathode using Al foil as the in-situ reductant. ACS Sustain. Chem. Eng. 2019, 14, 12222. [CrossRef] 13. Zhang, Y.; Li, Q.; Liu, X.; Xu, B.; Yang, Y.; Jiang, T. A Thermodynamic Analysis on the Roasting of Pyrite. Minerals 2019, 9, 220. [CrossRef] 14. Zhang, M.; Zhu, G.; Zhao, Y.; Feng, X. A study of recovery of copper and cobalt from copper–cobalt oxide ores by ammonium salt roasting. Hydrometallurgy 2012, 129, 140–144. [CrossRef] 15. Ding, W.; Xiao, J.H.; Peng, Y.; Shen, S.Y.; Chen, T. Iron Extraction from Red Mud using Roasting with Sodium Salt. Miner. Process. Extr. Metall. Rev. 2019.[CrossRef] 16. Güntner, J.; Hammerschmidt, J. Sulphating roasting of copper-cobalt concentrates. J. South. Afr. Inst. Min. Metall. 2012, 112, 455–460. 17. Han, P.W.; Xiao, L.; Wang, Y.L.; Lu, Y.G.; Ye, S.F. Cobalt Recovery by the Chlorination-Volatilization Method. Metall. Mater. Trans. B 2019, 50, 1128–1133. [CrossRef] 18. Zhao, Q.; Xue, J.; Chen, W. A Novel Self-Magnetizing Roasting Process for Recovering Fe from Low-Grade Pyrite Cinder and Blast Furnace Sludge. Trans. Indian Inst. Metals 2019, 72, 2547–2556. [CrossRef] 19. Xiao, J.H.; Ding, W.; Peng, Y.; Wu, Q.; Chen, Z.Q.; Wang, Z.; Peng, T.F. Upgrading iron and removing phosphorus of high phosphorus Oolitic iron ore by segregation roasting with Calcium chloride and Calcium hypochlorite. J. Min. Metall. Sect. B Metall. 2019, 55, 305–314. [CrossRef] 20. Zhang, G.F.; Yang, Q.R.; Yang, Y.D.; Wu, P.; McLean, A. Recovery of iron from waste slag of pyrite processing using reduction roasting magnetic separation method. Can. Metall. Q. 2013, 52, 153–159. [CrossRef] 21. Zhou, Y.; Xu, P.; Cheng, H.; Liu, Q. Thermal phase transition of pyrite from coal. J. Therm. Anal. Calorim. 2018, 134, 2391–2396. [CrossRef] 22. Zhang, G.F.; Shi, Z.; Yan, P.; Gao, L. Study on separation of iron from waste slag of pyrite processing. Sep. Sci. Technol. 2015, 50, 292–298. [CrossRef] 23. Jin, J.; Han, Y.; Li, H.; Huai, Y.; Peng, Y.; Gu, X.; Yang, W. Mineral phase and structure changes during roasting of fine-grained carbonaceous gold ores and their effects on gold leaching efficiency. Chin. J. Chem. Eng. 2019, 27, 1184–1190. [CrossRef] 24. Zhang, X.; Liu, F.; Xue, X.; Jiang, T. Effects of microwave and conventional blank roasting on oxidation behavior, microstructure and surface morphology of vanadium slag with high chromium content. J. Alloys Compd. 2016, 686, 356–365. [CrossRef] 25. Ilyas, S.; Srivastava, R.R.; Kim, H.; Ilyas, N.; Sattar, R. Extraction of nickel and cobalt from a laterite ore using the carbothermic reduction roasting-ammoniacal leaching process. Sep. Purif. Technol. 2020, 32, 115971. [CrossRef]

© 2020 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/).