Joint processing of V-bearing steelmaking slag and V-bearing black shale for vanadium and iron separation

To recycle vanadium from V-bearing steelmaking slag and V-bearing black shale, the both were jointly roasted to generate vanadium-rich phase, and then vanadium was separated by magnetic separation in this study. The compositions of samples were determined by X-ray fluorescence meter and the phases in the samples were characterized using X-ray diffractometer. Experimental results showed that, with increasing the ratio of CaO content to SiO 2 content in the samples, the vanadium separation efficiency first decreased, increased, and then decreased again. With increasing roasting temperature from 1423 to 1623 K, the vanadium separation efficiency increased. With increasing roasting time from one hour to four hours, the vanadium separation efficiency increased. The optimum conditions for vanadium recycling are the basicity of 1.2, roasting temperature of 1623 K, and roasting time of 4 hours. Under this condition, separation efficiency of vanadium reaches 71.6%, and the

as the shale to slag ratio and roasting temperature. V-bearing black shale from Hunan Province, China and V-bearing steelmaking slag from Maanshan Iron & Steel Company, Ltd. (Maanshan, China) were ground for 2 h in a ball grinder. The obtained shale powder was dried at 383 K for 12 h in a drying oven.
The steelmaking slag to black shale ratio was determined based on the metallurgical basicity principle (R), i.e., the ratio of the CaO content to the SiO2 content in the sample. The molar ratio of ferric oxide to carbon was fixed at 0.4 by adding graphite. Dried shale powder, slag powder, and graphite powder were mixed for 2 h in predetermined proportions in a blender mixer. The mixtures in industrial pure iron crucibles were placed in a sealed resistance furnace [36] and heated to the predetermined temperature at a rate of 5 K/min, roasted for a specified time at the temperature, cooled to room temperature at a rate of 5 K/min, and then the samples were removed from the furnace. Argon flow of 100 ml/min was maintained throughout the experiments. The experimental conditions for all of the samples are given in Table 1.
The roasted samples were ground into powder in a ball grinder. The powders were made into pulps by mixing them with tap water at a mass to volume ratio of 1:5. A magnetic tube (tube diameter: 50mm, head gap: 50mm, vibration frequency: 70times/min, magnetic field strength: 0-200KA/m) was used to process the pulps to obtain magnetic concentrates and tailings (magnetic intensity of 100KA/m was adopted). The obtained concentrates and tailings were dried at 383 K for 12 h in a drying oven.
The carbon content in the black shale was determined by loss on ignition. After grinding, the size distributions of the dried black shale powder, V-bearing steelmaking slag powder, and roasted samples were determined by the screening method. The chemical compositions of the dried black shale, V-bearing steelmaking slag, and dried concentrates and tailings were determined with an X-Ray fluorescence meter (ARL Advant' X Intellipower 3600) . The phase compositions of the dried black shale, V-bearing steelmaking slag, and roasted samples were determined by X-ray diffractometry (Bruker D8 Advance X-ray diffractometer, Cu K).
Based on the compositions of the concentrates and tailings, the vanadium and iron separation efficiencies were calculated by the Hancock efficiency equation: where E is the vanadium or iron separation efficiency (%), α is the mass percentage of V2O5 or Fe2O3 in the roasted sample, β is the mass percentage of V2O5 or Fe2O3 in the concentrate (%), θ is the mass percentage of V2O5 or Fe2O3 in the tailings (%), βi is the ideal mass percentage of V2O5 or Fe2O3 in the concentrate (%), and θi is the ideal mass percentage of V2O5 or Fe2O3 in the tailings (%). Based on the assumption that all of the vanadium and iron enter the vanadium-iron spinel, βi was obtained by calculation and θi was considered to be zero. The chemical compositions of the black shale and the steelmaking slag are given in Table 2. The black shale mainly consisted of silica and alumina. The steelmaking slag was composed of calcium oxide, iron oxide, silica, and magnesium oxide. The XRD patterns of the shale and the slag are shown in Figure 1. The shale mainly consisted of mica and silica. The slag was composed of tricalcium silicate, dicalcium silicate, calcium oxide, RO and dicalcium ferrite. Besides the aforementioned phases, according to our previous studies, the slag contained a supercooled glass. The glass phase formed from molten slag without crystallization, consisting of CaO, SiO2, FeOx and MgO. According to the results of X-ray diffractometry for the samples roasted at 1573 K , the pattern of the sample with a basicity of 0.6 is closely to that of the sample with a basicity of 0.7, the patterns of the samples with basicity of 0.8, 0.9, 1.0, 1.2 and 1.5 are similar, and the pattern of the sample with 1.7 is quite different from the others. Thus, the XRD patterns of the samples can be split into three groups. Three representatives from the three groups are shown in Figure 2. As seen in the figure, the samples with lower basicities (0.6 and 0.7) mainly consist of aluminosilicate similar to muscovite, which indicates original structure of mica isn't utterly destroyed, and the vanadium in the lattice of aluminosilicate isn't released in full. The sample with a basicity of nearly 1.0 contains Ca2MgSi2O7, KAlSiO4, Fe and Fe3O4 (V-bearing spinel phase). In the sample with a basicity of 1.7, two crystalline phases of dicalcium silicate and composite silicate occur. V-bearing spinel phase wasn't detected in the samples with low basicity and high basicity, which indicates that the formation of V-bearing spinel necessitates a moderate basicity.

Results
The particle size distribution of the shale powder was as follows: the 75-150, 48-75, and <48 μm fractions accounted for 11.6%, 41.8%, and 46.6% of the total mass, respectively. In the steelmaking slag powder obtained by grinding, the 75-150, 48-75, and <48 μm fractions accounted for 26.1%, 18.7%, and 55.2% of the total mass, respectively. Although the roasted samples were grounded for the same time, the particle size distributions were different because of the mineral composition difference. This difference was relative small, and the 75-150, 48-75, and <48 μm fractions accounted for 30%, 19%, and 51% of the total mass, respectively. The V2O5 and Fe2O3 contents in the concentrates and tailings, and the calculated separation efficiencies are given in Table 3. The results indicate that the basicity, roasting temperature, and roasting time have large effects on the vanadium and iron separation efficiencies from the roasted samples.

Theoretical analysis of V-concentrating phase formation
According to our previous study [36,37], the vanadium in the shale is mainly located in the mica, and the vanadium in steelmaking slag exists in a glass phase and dicalcium silicate. In the process of roasting, mica in the shale first broke down into silicate and water. The silicate and silica in turn reacted to calcium oxide, dicalcium silicate, tricalcium silicate, a glass phase mainly consisting of calcium ferrite and RO in the slag at higher temperature.

Fig. 3. Variation of standard Gibbs energy for several reactions with temperature
Standard Gibbs free energy for the possible reactions were plotted against temperature in Figure 3. The free energies are much less than zero, which indicates that the reactions may take place at 1273K to 1873K in term of thermodynamics. The reactions make the structures of mica, dicalcium silicate and the glass phase decompose. Vanadium oxide in the glass phase and crystal lattice of mica and dicalcium silicate was released, and iron oxide got liberated from glass phase and RO phase. In the study, the targeted compound of vanadium is V-bearing spinel (Fe3-mVmO4). The released vanadium oxide and iron oxide need further interact each other, forming V-bearing spinel (Fe3-mVmO4).
Vanadium and iron are the elements with variable valency and their valence is severely affected by oxygen partial pressure in atmosphere. Therefore, controlling oxygen partial pressure in atmosphere within an appropriate range is essential for obtaining the targeted compound. High pure argon gas can ensure the atmosphere [36].
Binary chemical basicity, i.e. the ratio of calcium oxide content to silica content, is another key factor influencing occurrence state of vanadium. With increasing the basicity, vanadium attempts to form high valence oxides, combining to calcium oxide. Our previous studies demonstrated that vanadium was contained in dicalcium silicate (Ca2Si1-xVxO4 or Ca2SiO4·nCa3V2O8) in the slags with high basicity [36]. Lower basicity improves the transform of iron element from high valence to low valence, reacting to acid oxides. In the studied system, FeSiO3 would be generated in substantial amounts, which restrains the generation of the spinel. Influences from the two aspects would leads to an optimal basicity.

Effect of the basicity
The effect of the basicity on the vanadium and iron separation efficiencies is shown in Figure 4. For the samples roasted at 1473 K, with increasing basicity, the V2O5 separation efficiency remained low (even less than 0), i.e., the V2O5 content in the concentrate was less than that in the tailings. The Fe2O3 separation efficiency decreased with increasing basicity, and it also maintained a low level. For the samples roasted at 1573 K, with increasing basicity, the V2O5 separation efficiency first decreased, increased, and then decreased, reaching a maximum when the basicity was 1.2. This are consistent with theoretical analysis of V-concentrating phase formation. The variation trend of the Fe2O3 separation efficiency was similar to that of the V2O5 separation efficiency. The Fe2O3 and V2O5 separation efficiencies are plotted against the roasting temperature in Figure 5. With increasing roasting temperature from 1423 to 1623 K, the Fe2O3 separation efficiency increased and then decreased, reaching a maximum at 1573 K. The V2O5 separation efficiency greatly increased with increasing roasting temperature from 1423 to 1573 K, and it slightly increased when the roasting temperature increased from 1573 to 1623 K.
The reasons for the above trends are as follows. On the one hand, high temperature favors formation and growth of the vanadium-iron spinel phase, which is favorable for vanadium and iron separation. On the other hand, higher temperature leads to more liquid phase forming, and more vanadium oxide and iron oxide dissolve in the liquid phase, which results in an increase in the iron oxide content in the tailings. The vanadium and iron separation efficiencies are plotted against the roasting time in Figure 6. With increasing roasting time from 1 to 3 h, the vanadium separation efficiency dramatically increased from 40.2% to 71.4%, and slightly increased to 71.6% when roasting time increased from 3 to 4 h. With increasing roasting time from 1 to 4 h, separation efficiency of iron gradually increased from 44.6% to 76.9%.

Effect of the roasting time
Generally, a long roasting time promotes crystal grain growth. Large crystal grains are conductive to separation, resulting in high separation efficiency. The experimental results are consistent with this principle.

Processing flow
Joint processing of V-bearing black shale and V-bearing steelmaking slag can be performed in two steps. In the first step, the mixture of the two is roasted. In the second step, after grinding, vanadium and iron are separated from the roasted material by magnetic separation, producing a concentrate and tailings.
The obtained concentrate mainly consisting of iron and vanadium can be used to produce V2O5 or vanadium-iron spinel. The tailings were mainly composed of calcium aluminosilicate and had high stability because of an appropriate basicity. Furthermore, the remaining vanadium in the tailings was present in the trivalent state in the form of spinel, which is relatively stable and non-toxic. Therefore, it can be used as building materials.

Conclusions
Based on investigation of joint roasting of V-bearing steelmaking slag and V-bearing black shale followed by magnetic separation, the effects of the roasting conditions on the vanadium and iron separation efficiencies can be summarized as follows.
1. With increasing basicity, in the samples roasted at 1573 K, the vanadium and iron separation efficiencies first decreased, increased, and then decreased, reaching a maximum when the basicity was 1.2. The vanadium and iron contents in the obtained concentrates showed a similar variation trend to the separation efficiencies. 2. Increasing the roasting temperature greatly affected the vanadium and iron separation efficiencies. The vanadium separation efficiency increased with increasing roasting temperature from 1423 to 1623 K. The iron separation efficiency increased with increasing roasting temperature from 1423 to 1573 K, and slightly decreased when the roasting temperature was increased from 1573 to 1623 K.
3. The roasting time had a significant effect on the vanadium and iron separation efficiencies. The vanadium separation efficiency increased from 40.2% to 71.4% when the roasting time was increased from 1 to 3 h, and it slightly increased from 71.4% to 71.6% when the roasting time was increased from 3 to 4 h. The iron separation efficiency increased from 44.6% to 76.9% when the roasting time was increased from 1 to 4 h. 4. Comprehensive evaluation of the vanadium and iron separation efficiencies and their contents in the concentrate showed that appropriate conditions are basicity of 1.2, roasting temperature of 1623 K, and roasting time of 4 h.

Figure captions
Fig. 1. X-ray diffraction patterns of black shale and steelmaking slag. Fig. 2. X-ray diffraction patterns of the samples roasted at 1573 K. Fig. 3. Variation of standard Gibbs energy for several reactions with temperature Fig. 4. Effect of the basicity on separation efficiencies of vanadium and iron Fig. 5. Effect of roasting temperature on separation efficiencies of vanadium and iron. Fig. 6. Effect of roasting time on separation efficiencies of vanadium and iron  Table 2 Chemical compositions of the black shale and steelmaking slag /mass % Table 3 Fe2O3 and V2O5 contents and separation efficiencies