ISIJ International
Online ISSN : 1347-5460
Print ISSN : 0915-1559
ISSN-L : 0915-1559
Fundamentals of High Temperature Processes
Crystallization Behavior of Synthesized CaO–SiO2–CaF2–La2O3 Rare Earth-containing Slag
Xuefeng She Zhenlong AnZhuo ZhangTengfei MaJingsong Wang
Author information
JOURNAL OPEN ACCESS FULL-TEXT HTML

2020 Volume 60 Issue 5 Pages 832-839

Details
Abstract

Crystallization behaviors of CaO–SiO2–CaF2–La2O3 synthesized slag with different basicity have been studied by differential scanning calorimetry (DSC), field emission scanning electron microscopy (SEM) and X-ray diffraction (XRD). It was found that there were mainly three kinds of crystalline phases (La7.58(Si1.048O4)6O2, CaSiO3 and Ca4Si2O7F2) precipitated in the slag during the cooling process. Crystalline phase Ca2SiO3 formed when the basicity was increased to 1.3. The observations confirmed that the rare earth phase (La7.58(Si1.048O4)6O2) precipitated firstly in each slag during the cooling process. With the increase of slag basicity, the crystallization temperature of the rare earth phase decreases while the precipitation peak temperature of CaSiO3 and Ca4Si2O7F2 increase. The morphology of rare earth phase is hollow hexagonal which is filled with substrate phase. The optimum condition for crystallization and separation of rare earth phase was obtained. The basicity of the slag should be controlled between 0.9 and 1.1. The morphology of rare earth phases can grow much better after isothermal heat treatment for 4 h at the crystallization temperature of rare earth phases. The mean size of rare earth phase could increase to more than 60 µm by isothermal heat treatment.

1. Introduction

Rare earth, an important raw material of national defense industry, is one of the most significant strategic resources related to national security and development.1) Despite the relatively abundant content of rare earth elements (RE) in the earth’s crust, the distribution of RE in the world is extremely uneven and the exploitable rare earth ores that have been discovered are relatively less than other minerals.2) It’s well known that the Bayan Obo Mine in the Inner Mongolia region of North China is the largest iron-LREE-niobium ore deposit in the world. It accounts for 35% rare-earth resource of world’s proven reserves and approximately 80% of China.3,4,5) Considering that the designed capacity of the Bayan Obo ore is 12 million tons per year, the rest ore will be used up within 20 years. Therefore, it’s an urgent task to make Bayan Obo ore be utilized comprehensively.6,7)

RE-bearing slag is acquired from Bayan Obo ore by direct reduction and melting process. However, the rare earth elements content is low in the slag. It is well known that the integrity of mineral grain is conducive to crushing, grinding and separation for beneficiation process. Thus it is crucial to study the law of RE phase crystallization for the purpose of enriching the RE elements and making the RE-phase growth well. Over the past decades, many researchers have studied the properties of the RE- bearing slag. Li et al.8,9) studied the crystallization behavior of cerite calcium phase in blast furnace slag in isothermal process by quenching method, and they found the nucleation of cerite calcium phase is dominated by homogeneous nucleation. Hao et al.10,11) researched the crystallization process of rare earth slag by putting slag into a special heat-retarder, and they found that the grains of different minerals in the slag are fully developed and grown. Ding et al.12,13) studied the leaching behavior of rare earth slag produced by carbon-bearing pellet reduction and melting process, which is beneficial for the study of RE-bearing slag. Guo et al.14,15,16) reported that the rare earth phase could be separated from slag by super-gravity. They have made some progress and the recovery ratio of RE in the concentrate is up to 29.96 pct. However, the extraction of RE element from molten slag by these methods has a high cost and a great amount of acid was consumed.

In order to enrich the dispersed target elements into a selected crystalline phase, selective crystallization and phase separation (SCPS) method was proposed. The method has been widely studied to deal with the boron-bearing slags, vanadium-bearing slags and the titanium-bearing slags.17,18,19) So it can also be used to recover RE from the RE-bearing slag. By adjusting the composition of the RE-bearing slag and controlling the cooling process, the dispersed RE elements are enriched to a certain phase and the RE-enriched crystals grows to a large enough size. However, it hasn’t been investigated that how the composition of the slag affect the crystallization behavior of the RE-bearing slag. Besides, there are no isothermal heat treatment has been found for rare earth slag in previous study.

This work is further studied based on Ding’s experiment.12) In the simulation test of rotary hearth furnace conducted using Bayan Obo complex ore, the main composition of acquired slag are RE2O3 (14.19%), CaF2 (36.82%), CaO (18.19%) and SiO2 (18.01%). Meanwhile, in order to avoid the influence of cerium with different valence states on slag cooling, lanthanum was selected as the most primary research object in this study. So the La2O3–CaO–SiO2–CaF2 system is considered as a mother phase for the slag. The differential scanning calorimetry (DSC) has been employed to explore the influence of basicity and cooling process on the crystallization behavior of the RE-bearing slag. The crystalline phases in the RE-bearing slag were identified by XRD. The scanning electron microscope (SEM) equipped with energy dispersive spectroscopy (EDS) were applied to determine the size, morphology and composition of crystals. This work has the potential to provide reference for the recover of rare earth elements in the beneficiation process.

2. Experimental and Methods

2.1. Sample Preparation

In order to eliminate the impact of minor elements in the industrial RE-slag, such as K, Na, Fe and other elements, the synthesized slag was prepared using reagent grade CaO, SiO2, CaF2 and La2O3. These materials were weighed carefully to obtain the target composition shown in Table 1. The basicity varied from 0.7 to 1.3, and the content of La2O3 and CaF2 was fixed as 15% and 10%, respectively. In this study, glass sample of the quaternary CaO–SiO2–CaF2–La2O3 were prepared by the conventional melting-quenching method. The composition after pre-melting was determined via X-ray fluorescence (XRF), and the results (Table 1) show that there are small deviations from the target compositions.

Table 1. Target and actual compositions of the experimental slag samples (mass%).
slagTarget CompositionXRF Analysis Results
CaOSiO2La2O3CaF2BasicityCaOSiO2La2O3CaF2Basicity
130.9044.1015.0010.000.7032.1043.9814.319.610.73
235.5039.5015.0010.000.9036.3039.8814.259.570.91
339.3035.7015.0010.001.1037.0737.8215.459.660.98
442.4032.6015.0010.001.3044.6031.8614.119.431.40

Chemicals used in the experiment were thoroughly mixed in ethyl alcohol and dried at 378 K (105°C) for 12 h. Then 3 g of synthesized slag in a graphite crucible were melted under argon atmosphere at 1723 K (1450°C) for 2 h in order to completely homogenize the slag. After pre-melting, the sample was then quenched into water to produce the glassy slag. Subsequently, the sample was dried at 378 K (105°C), crushed and ground to 300 meshes for the further analysis by the differential scanning calorimetry (DSC). Furthermore, the composition of the sample was analysed by X-Ray fluoroscopy (XRF).

2.2. DSC Measurement

The crystallization behaviors of RE-slags were investigated by DSC. In the process of the DSC measurement, the sample was placed in a platinum crucible in the atmosphere of Ar with a flow rate of 60 mL/min. About 50 mg of slag powder was heated at a constant heating rate of 15 K/min from room temperature to 1723 K (1450°C), and then kept the temperature at 1723 K (1450°C) for 20 min to make sure the melt of the sample and homogenize its chemical composition. After that, the sample was cooled to room temperature, and the cooling rate was fixed at 5 K/min. The temperature schedule of the DSC measurement is shown in Fig. 1. The DSC date was recorded automatically and was further analyzed.

Fig. 1.

Thermal history in non-isothermal DSC measurements.

Due to the little amount of the sample measured by DSC, greater amount of slag was heated in a tube furnace at Ar atmosphere with the same temperature schedule as DSC measurement which in order to study the crystalline structure and morphology of the cooled slag by XRD and SEM. The slag was divided into two parts. One part was ground for XRD analysis and the other was mounted in epoxy resin, then polished and sprayed with a carbon film for cross section observation by SEM-EDS.

2.3. Non-isothermal Heat Treatment and Crystalline Phase Analysis

In order to determine the crystalline phase of the slag corresponding to each exothermic peak of DSC curves, a series of continuous cooling experiments were carried out in a tube furnace. As shown in Fig. 2, a MoSi2 resistance furnace, which the cooling speed can be controlled, was used to heat a corundum tube of 10 cm in the diameter and 105 cm in the height. In addition, the uniform temperature zone is 65 to 75 cm below the furnace mouth. Before the experiment, the temperature of the uniform temperature zone was raised to 1723 K (1450°C). About 5 g of slag were put into a graphite crucible hung on a molybdenum wire with a length of 70 cm, and then, the crucible was put into the furnace and maintained in the uniform temperature zone for 30 min. After being cooled to the target temperature with a constant speed of 5 K/min, the sample was quenched with ice water. The phase composition of the samples were determined by XRD. The morphology and elemental composition of different samples were observed employing a scanning electron microscopy (SEM) equipped with energy dispersive spectrometry (EDS). In this experiment, the target temperature was determined based on the exothermic peak of DSC curves as shown in Table 2.

Fig. 2.

Schematic diagram of the MoSi2 resistance furnace. (Online version in color.)

Table 2. Desired target temperature in non-isothermal heat treatment experiments.
Exp No.SampleTarget Temperature [K (°C)]
1Slag 21633 (1360)
2Slag 21448 (1175)
3Slag 21438 (1165)
4Slag 31466 (1193)
5Slag 31573 (1300)
6Slag 41545 (1272)
7Slag 41491 (1218)
8Slag 41456 (1183)

3. Results and Discussion

3.1. Non- isothermal DSC Measurement and XRD Identification

Figure 3 shows the results of DSC measurement for glass samples with different basicity. The exothermic peaks attributed to crystallization of the slag appeared between 1150°C and 1400°C. The crystallization temperatures and crystal formation events of the slag can be determined from the exothermic peaks in the DSC curves.

Fig. 3.

DSC curves of non-isothermal crystallization of slags at a cooling rate of 5 K/min. (Online version in color.)

In Fig. 3, there is no exothermic peak on DSC curve of slag 1, whereas two or three exothermic peaks can be found for slag 2, 3 and 4. These results indicate that basicity has a great influence on the crystallization of RE-bearing slag. With the increase of basicity, different crystalline phases are produced. CaO, as a basic oxide, can depolymerize the complex Si–O network structure. The viscosity of the slag decreases and mass transfer becomes more flexible with the increase of basicity,20,21,22) thus the crystallization of the slag is promoted. For slag 1, the kinetics of crystallization are poor due to the high viscosity of slag at low basicity. It is believed that the phases in the slag 1 generated so slowly that the heat release caused by crystallization could not be detected by DSC measurement at the cooling rate of 5 K/min.

In order to further investigate the properties of the RE-bearing slag with different basicity, the slags cooled with the same temperature schedule as DSC measurement were also studied experimentally. Due to the lack of relevant date about pure RE-bearing crystalline phase, the following RE-bearing crystalline phase was considered as La7.58(Si1.048O4)6O2, which characteristic peak is basically consistence with the results of XRD.

Figure 4 shows the XRD patterns of the slags with different basicity. There are only two crystalline phases, which are CaSiO3 and La7.58(Si1.048O4)6O2, precipitated in the continuous cooling process for slag 1, whereas three crystalline phases, CaSiO3, Ca4Si2O7F2 and La7.58(Si1.048O4)6O2 can be found for slag 2 and slag 3. As for slag 4, four crystalline phases, CaSiO3, Ca4Si2O7F2, Ca2SiO3 and La7.58(Si1.048O4)6O2, are detected. These results indicate that the crystallization behavior of RE-bearing slag is strongly affected by the increase of basicity. There is a competition between cuspidine and wollastonite formation in the RE-bearing slag. The formation of cuspidine (Ca4Si2O7F2) is suppressed in the slag with the basicity of 0.7, which is the reason that no cuspidine (Ca4Si2O7F2) is found in XRD pattern. These results are consistent with the studies of Seo et al.23) With the increase of basicity, cuspidine precipitated in the slag 2, slag 3 and slag 4. On the one hand, there is more CaO remained in the slag for the crystallization of cuspidine. On the other hand, the kinetics for crystallization is improved with the increase of basicity. With the further increase of CaO content, dicalcium silicate generated when the basicity of the slag was increased to 1.3.

Fig. 4.

XRD pattern of the slag cooled with the same temperature schedule as DSC measurement. (Online version in color.)

3.2. Observation of Microstructure of the Rare Earth Slag

The microstructure of the slag cooled with the same temperature schedule as DSC measurement is shown in Fig. 5. According to the results of both XRD pattern and EDS analysis, the rare earth phase (La7.58(Si1.048O4)6O2) exists in the bright white part while cuspidine (Ca4Si2O7F2) and wollastonite (CaSiO3) present as grey and dark grey, respectively. Figure 5(a) shows that there are two crystalline phases, rare earth phase (La7.58(Si1.048O4)6O2) and wollastonite (CaSiO3), precipitated in the slag. However, there are no exothermic peaks on DSC curve at the cooling rate of 5 K/min. This is a result of low heat liberation effect of the slag in the crystallization process. At the basicity of 0.7, high content of SiO2 in slag can cause high slag viscosity and inferior slag fluidity, which increase the transfer resistance of ion clusters and the energy bar needed for the nucleation and growth of crystal.24,25) Therefore, the heat flux of crystallization was too weak to be detected by DSC measurement.

Fig. 5.

Microstructure of the slag cooled with the same temperature schedule as DSC measurement. (a) slag 1, (b) slag 2, (c) slag 3, (d) slag 4. (Online version in color.)

As shown in Fig. 5(b), the rare earth phase (La7.58(Si1.048O4)6O2) is in an irregular shape and randomly distribute in the slag. This is due to the fact that crystallization of rare earth phase is not complete as a result of poor crystallization kinetic conditions. The black grey rod-like crystal is wollastonite (CaSiO3), and the size of which is very large. This indicates that the crystallization of wollastonite has an absolute advantage in the slag at low basicity, although it is not the first precipitated phase. There is a reason to believe that wollastonite (CaSiO3) can crystallize on the basis of Si–O network structure. Therefore, wollastonite phase (CaSiO3) gets good growth while the crystallizations of the other phases are inhibited. Unlike the XRD pattern, no cuspidine (Ca4Si2O7F2) is found as a result of its little amount.

Figure 5(c) shows the microstructure of the slag at the basicity of 1.1. The distribution of rare earth phase (La7.58(Si1.048O4)6O2) becomes more concentrated with the increase of basicity. Still, the crystallization of rare earth phase have not fully developed and the morphology of which is chicken claw in shape. The large rod-like wollastonite (CaSiO3) disappeared and it was replaced by smaller dendritic one. And cuspidine (Ca4Si2O7F2) with no fixed shape, as a substrate phase, surrounds the wollastonite (CaSiO3). It is due to the depolymerization of Si–O reticular formation at higher basicity.

As shown in Fig. 5(d), the size of rare earth phase increases with the further aggregation of the rare earth phase. The morphology of rare earth phase is hollow hexagonal which is filled with substrate phase. Compared to Figs. 5(a), 5(b), 5(c), the rare earth phase gradually grew from a fine dispersing floc or spicule to a regular hexagon with a diameter of 60 μm, which indicates that the crystallization of rare earth phase has developed well. Substrate phase is dominated by dicalcium silicate (Ca2SiO3), in which cuspidine and wollastonite are scattered. It is noteworthy that there are incomplete hexagonal rare earth phases in the slag, which indicates the crystallization of rare earth phase could still be enhanced. Therefore, in section 3.4, isothermal heat treatment under crystallization temperature of rare earth phase was carried out to make the crystallization of rare earth phase fully developed.

3.3. Identification of Crystalline Phases

The slag was maintained in the even temperature zone (1773 K) for 30 min to ensure complete melt and homogenization. This was followed by a continuous cooling to the target temperature as listed in Table 2. After 30 minutes’ heat preservation at the target temperature, the sample was quenched into water. The XRD and SEM were employed to determine the crystal phases corresponding to the exothermic peaks on DSC curves. Figure 6 displays the XRD patterns and phase identification of the slags quenched at desired target temperature corresponding to these exothermal peaks of DSC curves. The results of phase identification are summarized in Table 3. As shown in Fig. 6(a), the XRD analysis reveals that the first crystalline phase in slag 2 (P1) corresponds to rare earth phase (La 7.58(Si1.048O4)6O2). It could be confirmed by XRD that two new phases generate at the second exothermal peak P2. As shown in Fig. 3, the two exothermal peaks of P2 and P3 for slag 2 are close together, which indicates that crystals represented by P2 and P3 can precipitate at almost the same temperature. The XRD patterns of P2 and P3 are similar, which in agreement with the DSC result. As for slag 3, it could be noted from Fig. 6(b) that the first exothermal peak (P1) represents the formation of rare earth phase (La7.58(Si1.048O4)6O2). Both cuspidine (Ca4Si2O7F2) and wollastonite (CaSiO3) were detected by XRD at the second peak of slag 3. Therefore, it could be inferred that cuspidine (Ca4Si2O7F2) and wollastonite (CaSiO3) crystalline simultaneously. As shown in Fig. 6(c), three crystalline phases appeared due to the strong crystallization ability when the basicity reaches 1.3. Based on previous results of slag 2 and slag 3, it can be deduced that the first exothermal peak (P1) on the slag 4 is rare earth phase (La7.58(Si1.048O4)6O2). Dicalcium silicate (Ca2SiO3) and cuspidine (Ca4Si2O7F2) then appeared at P2 and P3.

Fig. 6.

XRD pattern of the slag quenched at desired target temperatures: (a) slag 2, (b) slag 3, (c) slag 4. (Online version in color.)

Table 3. XRD analysis result of crystalline phase in the slag quenched at desired target temperatures.
SampleTarget Temperature [K (°C)]Crystalline Phase Identified by XRD
Slag 21633 (1360)La 7.58(Si1.048O4)6O2
Slag 21448 (1175)La 7.58(Si1.048O4)6O2 + CaSiO3 + Ca4Si2O7F2
Slag 21438 (1165)La 7.58(Si1.048O4)6O2 + CaSiO3 + Ca4Si2O7F2
Slag 31573 (1300)La 7.58(Si1.048O4)6O2
Slag 31466 (1193)La 7.58(Si1.048O4)6O2 + CaSiO3 + Ca4Si2O7F2
Slag 41545 (1272)La 7.58(Si1.048O4)6O2 + CaSiO3 + Ca4Si2O7F2
Slag 41491 (1218)La 7.58(Si1.048O4)6O2 + CaSiO3 + Ca4Si2O7F2
Slag 41456 (1183)La 7.58(Si1.048O4)6O2 + CaSiO3 + Ca4Si2O7F2

Microstructure of the slag 2, slag 3 and slag 4 quenched at the first exothermal peak P1 are shown in Fig. 7. It confirms that the rare earth phase crystallized at the first exothermal peak P1 in the crystallization process of slag 2, slag 3 and slag 4, by combining EDS results with XRD patterns. Other crystalline phases like cuspidine (Ca4Si2O7F2) and wollastonite (CaSiO3) hardly find in Fig. 7, for the quenching temperature is much higher than the crystallization temperature of them.

Fig. 7.

Microstructure of the slag quenched at the first exothermal peak P1: (a) slag 2, (b) slag 3, (c) slag 4. (Online version in color.)

3.4. Effect of the Isothermal Heat Treatment on Crystallization of Rare Earth Phase

In order to study the effect of cooling regime on the crystallization and growth of rare earth phase, the isothermal treatment experiment was carried out. The samples, chosen as slag 2 and slag 3, were melted in a muffle furnace at 1773 K (1500°C) for 30 min, and then cooled rapidly to the crystallization temperature of the rare earth phase and kept for 2 h and 4 h at that temperature (1633 K for slag 2 and 1573 K for slag 3). After that, the slags were taken out and rapidly cooled down to room temperature in the N2 flow.

Figure 8 shows the microstructure of the slag 2 after isothermal heat treatment. When the holding time is 2 h, it can be seen from Fig. 8(a) that the main crystalline phase in the slag is calcium silicate (CaSiO3), which is in long strip shape. Cuspidine (Ca4Si2O7F2), as the substrate phase, fills the part outside the REE-producing phase and the calcium silicate phase. As shown in Fig. 8(b), most of rare earth phases have no fixed shape and the crystal size is small. In addition, the finely dispersed rare earth phases are scattered in the substrate phase. These indicate that the rare earth phase has not fully grown up yet, and it is very difficult to separate the rare earth phase in this state.

Fig. 8.

Microstructure of the slag with the basicity of 0.9 after isothermal heat treatment. (a), (b) and (c) 2 hours’ isothermal heat treatment; (d), (e) and (f) 4 hours’ isothermal heat treatment. (Online version in color.)

After holding for 4 h, it can be seen from Fig. 8(d) that the size of rare earth phase increase obviously and most of the rare earth phase exists in the form of hexagon. Besides, there is basically no fine rare earth phase distributed in the substrate phase. Compared with Fig. 8(a), there is no obvious change in calcium silicate crystal except further aggregation of finer calcium silicate crystalline phases. With the magnification of SEM increased to 400 times, Fig. 8(e) shows that the cross section of rare earth phase is in a regular hexagon with impurities in the center. The cross-section area of most rare earth phases is between 80 μm and 100 μm, and some of them exceed 100 μm.

The microstructure of the slag 3 after isothermal heating treatment is shown in Fig. 9. When the holding time is 2 h, most of the rare earth phase is needle-like and the rest exists in the form of irregular hexagon. Compared with slag 2, rare earth phases are more concentrated and there is almost no fine dispersed rare earth phase in the substrate phase. When the holding time is prolonged to 4 h, the crystallization of rare earth phase is further developed, the shape of which is more regular and the size is also increased. The size of most of rare earth phases is between 60 μm and 80 μm, which is a little smaller than that in slag 2.

Fig. 9.

Microstructure of the slag with the basicity of 1.1 after isothermal heat treatment. (a), (b) and (c) 2 hours’ isothermal heat treatment; (d), (e) and (f) 4 hours’ isothermal heat treatment. (Online version in color.)

Compared with the microstructure of the slag quenched at the target temperature which is shown in Fig. 7, it can be seen that the size of rare earth phase can significantly increase and the crystal are further developed by heat holding treatment. According to classical nucleation theory, nucleation rate has a maximum value.26,27) The driving force of nucleation increases with the increase of undercooling in view of thermodynamics.28) However, mass transfer resistance will be higher with the increases of viscosity, which counterbalances the thermodynamic factors and reduces the nucleation rate. There is a similar relationship for crystal growth.29) Therefore, nucleation rate and growth rate at different undercooling are shown in Fig. 10. When the slag is kept at a higher temperature in the crystallization zone, the RE-crystal size tends to be larger and the number of RE-crystals tends to be less due to the slower nucleation rate and the faster growth rate of the rare earth phase. In addition, there is more time for mass transfer in the process of heat holding. The phenomenon of crystal growth during heat preservation was mentioned in many other studies.30,31,32)

Fig. 10.

Nucleation rate and growth rate at different undercooling. (Online version in color.)

The size of rare earth phase decreases a little with the basicity increases from 0.9 to 1.1. As shown in Fig. 11, at the basicity of 0.9, there are three phases, La 7.58(Si1.048O4)6O2, CaSiO3 and Ca4Si2O7F2. The content of rare earth elements is roughly equal to that of CaSiO3. Whereas, when the basicity increase to 1.1, the RE elements’ content is higher than that of CaSiO3. Thus, there is a competition between rare earth phase and wollastonite formation in the slag with higher basicity. The increase of CaO content enhances the binding ability of CaO to SiO2, as a result of which the combination of La2O3 and SiO2 is inhibited. Therefore, according to the DSC results of Fig. 3, the crystallization peak temperature of the rare earth phase decreases and the exothermic peak intensity weakens while the crystallization peak temperature of wollastonite and cuspidine increase. However, the increase of basicity is conducive to crystallization from the aspect of kinetics. The degree of polymerization (DOP) of the structure decreases with the increase of CaO content, which improves the fluidity of the slag. The crystallization is promoted due to the reduction of mass transfer resistance. This also explains why there are three phases for the slag 4 after quenching at 1545 K. Therefore, basicity of the slag should be suitable in order to promote the enrichment of rare earth elements. In this experiment, basicity from 0.9 to 1.1 is more appropriate for beneficiation process.

Fig. 11.

Compositions of crystalline phase at different basicity. (Online version in color.)

4. Conclusions

Crystallization Behaviors of synthesized CaO–SiO2–CaF2–La2O3 slag with various basicity have been studied by differential scanning calorimetry (DSC), XRD, SEM-EDS. Based on the results of this study, the following conclusions are presented:

(1) There are mainly three kinds of crystalline phases, La7.58(Si1.048O4)6O2, CaSiO3 and Ca4Si2O7F2 precipitated in the slag during the cooling process. When the basicity is increased to 1.3, new crystalline phase Ca2SiO3 formed.

(2) The morphology of rare earth phase is hollow hexagonal which is filled with substrate phase. The precipitation temperatures of rare earth phases at different basicity are: 1633 K for slag 2, 1573 K for slag 3 and 1545 K for slag 4.

(3) The rare earth phase is precipitated firstly during the cooling process. With the increase of basicity, the crystallization temperature of the rare earth phase decreases while the precipitation peak temperature of wollastonite and cuspidine increase.

(4) The optimum condition for crystallization and separation of rare earth phase as follows: the basicity between 0.9 and 1.1, and isothermal heat treatment for 4 h at the crystallization temperature of rare earth phases.

(5) By isothermal heat treatment, the size of rare earth phase could be increased to more than 60 μm which meets the mineral processing requirements of rare earth slag.

Acknowledgements

This work is supported by the The National Natural Science Foundation of China (51874029), the projects of State Key Laboratory of Advanced Metallurgy (41617015) and (41618018), which is acknowledged with thanks.

References
 
© 2020 by The Iron and Steel Institute of Japan
feedback
Top