Phase Equilibrium Investigation for CaO-SiO₂-5wt.%MgO-20wt.%Al₂O₃-TiO₂ System Relevant to Ti-bearing Slag System

Abstract

In order to resolve the lack of thermodynamic information for the comprehensive unilization of Ti-bearing blast furnace slag, the phase diagram relevant to the Ti-bearing slag composition was widely investigated. In the present work, the phase equilibrium relationships were investigated for the CaO-SiO₂-5wt.%MgO-20wt.%Al₂O₃-TiO₂ phase diagram system. The equilibrium phases were experimentally determined at 1300°C and 1400°C using the high temperature equilibrium technique followed by X-Ray Fluorescence (XRF), X-Ray Diffraction (XRD), Scanning Electron Microscope (SEM) and Energy Dispersive X-ray spectroscope (EDX) analysis. The liquid phase (L), melilite solid solution ((2CaO·MgO·2SiO₂, 2CaO·Al₂O₃·SiO₂)_ss) phase, perovskite (CaO·TiO₂) phase, Al–Ti diopside phase and pseudobrookite solid solution (MgO·TiO₂, Al₂O₃·TiO₂)_ss phase were found. Based on the experimental results, the 1300°C to 1500°C liquidus lines and phase diagram were constructed for the specified region of the CaO-SiO₂-5wt.%MgO-20wt.%Al₂O₃-TiO₂ system.

1. Introduction

With the decreasing of high grade ore resources, the comprehensive utilization of secondary resources has become more and more important all over the world. It is reported that about 3 million tons of Ti-bearing blast furnace slag, which contains more than 20 wt.% TiO₂,¹⁾ is produced every year in China. Ti-bearing slag can be regarded as the CaO–SiO₂–MgO–Al₂O₃–TiO₂ system, and there is no doubt that Ti-bearing slag can be taken as a valuable secondary resource. Until now, many processes have been proposed for treating the slag, such as the acid leading,^2,3) carbide-chloride method^4,5,6) and the selective crystallization and phase separation method (SCPS).^7,8) SCPS method has been considered as a promising method to dispose the slags and the key of SCPS method is to enrich the Ti element into a certain phase, such as perovskite,⁹⁾ rutile^10,11) and anosovite¹²⁾ and to make the selected phase to fully grow. However, the lack of thermodynamic information especially the phase diagram data for Ti-bearing furnace slag severely restricts the development of above mentioned processes, hence, the determination of the phase relations and liquidus temperatures for Ti-bearing slag system becomes more and more important for the comprehensive utilization of the Ti resources.

The phase relations of Ti-bearing slag system have been extensively investigated in the past few decades. As the most basic system of Ti-bearing slag, the phase diagram of CaO–SiO₂–TiO₂ system was completely investigated by DeVries et al.¹³⁾ and the addition of 10 wt.%–20 wt.% Al₂O₃ to CaO–SiO₂–TiO₂ system was studied by Ohno et al.¹⁴⁾ in air. It was found by Ohno et al.¹⁴⁾ that the liquidus temperature was decreased with the increase of TiO₂, while it was reported by Fine et al.¹⁵⁾ that the addition of TiO₂ increases the liquidus temperature. Osborn et al.¹⁶⁾ studied the effect of TiO₂ on liquidus temperatures in CaO–SiO₂–MgO–Al₂O₃ system, and the results indicated that the liquidus temperatures were not greatly influenced.¹⁷⁾ In order to find out the primary crystallization region of anosovite phase, Wang et al.¹⁸⁾ studied the phase equilibria in the TiO₂ rich part of the TiO₂-CaO-SiO₂-10wt.%Al₂O₃ system at 1500°C and 1600°C, and it was reported that the addition of 10 wt.% Al₂O₃ into TiO₂–CaO–SiO₂ system enlarges the low liquidus temperature regions, but the precipitation of anosovite phase was not found in the experiments by Wang et al.¹⁸⁾ The phase equilibrium at reducing condition was experimentally studied by Zhao et al.¹⁹⁾ and the melting temperature was described as a function of basicity and TiO_x concentration, and the possibility of recovering titania in the form of pseudobrookite was also suggested by Zhao et al.¹⁹⁾ However, the phase relations for CaO–SiO₂–MgO–Al₂O₃–TiO₂ system still remain unclear and the update of the fundamental thermodynamic information^20,21,22) becomes extremely urgent for the development and application of aforementioned system.

Therefore, a series of researches^{23,24,25,26,27)} has been proposed for the phase diagram investigation for CaO–SiO₂–MgO–Al₂O₃–TiO₂ system. In the work by Shi et al.^23,27) and Sun et al.,²⁵⁾ the 1300°C–1500°C liquidus lines for 10 wt.% and 20 wt.% Al₂O₃ section in CaO-SiO₂-5wt.%MgO-Al₂O₃-TiO₂ system have been established by SHTT (Single Hot thermocouple Technique),²⁸⁾ which can realize in-situ accurate and reliable determination for the liquidus temperature of high melting points slags.^29,30) Based on these results, the equilibrium phase relations are further determined for 20 wt.% Al₂O₃ section in CaO-SiO₂-5wt.%MgO-Al₂O₃-TiO₂ system. The results from present work and literature data are combined together to construct a phase diagram for the specific region in the CaO-SiO₂-5wt.%MgO-20wt.%Al₂O₃-TiO₂ system.

2. Experimental Procedure

2.1. Sample Preparation

The compositions used for equilibrium experiments were selected in reference to the 1300°C and 1400°C isotherms by Sun et al.,²⁵⁾ as shown in Figs. 1(a) and 1(b) and will be introduced in detail in the following section. Reagent grade oxides powders (Sinopharm Chemical Reagent Co., Ltd, China) of CaO (99.99 wt pct pure), SiO₂ (99.99 wt pct pure), MgO (99.99 wt pct pure), Al₂O₃ (99.99 wt pct pure) and TiO₂ (99.99 wt pct pure) were employed to synthesize the slags, which were calcined at 1000°C for 4 h to evaporate the moisture and impurities, carefully weighted (to an accuracy of 0.1 mg), fully mixed and pre-melted in Ar atmosphere by a vertical MoSi₂ furnace with temperature accuracy estimated to be ±2°C. The mixtures were placed in platinum crucibles which were suspended by platinum wire inside the hot zone of the furnace at 1650°C for 2 h to completely homogenize the slag. The samples were then quenched into ice water, and the X-Ray Powder Diffraction (XRD, X’Pert Pro, Panalytical B V, with Cu Kα radiation) analyses, metallographic microscope analysis and Scanning Electron Microscope (SEM, JXA-8530F, JEOL) analysis proved that the quenched slags showed glassy phase. The generator voltage and tube current of XRD were 40 kV and 40 mA, respectively. The scanning range from 5°–80° was measured at a step of 0.02°. The quenched samples were then dried, crushed and ground under 200 mesh for further utilization. The composition of quenched slags may differ from those of the designed samples because of evaporation losses, like of H₂O from CaO absorbed after calcining. X-Ray Fluorescence (XRF, ZSX100e, Rigaku) was then applied to analyze the compositions for comparison. The results of both are listed in Table 1, it can be seen that the measured values showed a small deviation compared with the designed composition. In order to visualize the distribution of the pre-melted samples, the measured compositions are projected on the CaO-SiO₂-5wt.%MgO-20wt.%Al₂O₃-TiO₂ pseudo-ternary phase diagram, as shown in Figs. 1(a) and 1(b). In Fig. 1(b), the 1300°C and 1400°C liquidus from Sun et al.²⁵⁾ were also plotted for comparison.

Fig. 1.

Projection of the pre-melted compositions on CaO-SiO₂-5wt.%MgO-20wt.%Al₂O₃-TiO₂ phase diagram, wt.%.

Table 1. Comparison of designed and measured compositions for Ti-bearing slag systems.

Sample No.		Compositions, wt.%
		TiO2	SiO2	CaO	MgO	Al2O3	Total
K1	Measured	5.33	35.45	32.62	5.96	20.64	100
	Designed	5.00	38.89	31.11	5.00	20.00	100
K2	Measured	5.01	33.00	36.29	5.90	19.80	100
	Designed	5.00	35.00	35.00	5.00	20.00	100
K3	Measured	5.18	29.47	40.07	6.04	19.24	100
	Designed	5.00	31.82	38.18	5.00	20.00	100
K4	Measured	4.54	28.11	42.72	5.29	19.34	100
	Designed	5.00	28.00	42.00	5.00	20.00	100
K5	Measured	4.97	26.97	43.15	5.91	19.00	100
	Designed	5.00	28.00	42.00	5.00	20.00	100
K6	Measured	9.67	30.61	35.20	5.63	18.89	100
	Designed	10.00	32.50	32.50	5.00	20.00	100
K7	Measured	10.05	27.65	38.44	5.58	18.28	100
	Designed	10.00	29.55	35.45	5.00	20.00	100
K8	Measured	9.58	25.19	41.31	5.06	18.86	100
	Designed	10.00	26.00	39.00	5.00	20.00	100
K9	Measured	15.13	30.90	30.28	5.15	18.54	100
	Designed	15.00	33.33	26.67	5.00	20.00	100
K10	Measured	14.97	25.27	36.64	5.01	18.11	100
	Designed	15.00	26.09	33.91	5.00	20.00	100
K11	Measured	14.23	23.36	38.78	4.44	19.19	100
	Designed	15.00	24.00	36.00	5.00	20.00	100
K12	Measured	20.81	24.19	30.23	4.92	19.85	100
	Designed	20.00	27.50	27.50	5.00	20.00	100
K13	Measured	18.65	22.64	35.78	4.58	18.35	100
	Designed	20.00	22.18	32.82	5.00	20.00	100
K14	Measured	20.74	19.97	35.74	4.33	19.22	100
	Designed	20.00	22.00	33.00	5.00	20.00	100
K15	Measured	28.28	23.21	25.30	4.43	18.78	100
	Designed	25.00	27.78	22.22	5.00	20.00	100
K16	Measured	25.64	22.53	28.38	4.79	18.66	100
	Designed	25.00	25.00	25.00	5.00	20.00	100
K17	Measured	25.83	20.10	30.54	4.57	18.96	100
	Designed	25.00	22.73	27.27	5.00	20.00	100

2.2. Equilibration Experiments

The vertical furnace used for pre-melting was employed for the equilibrium experiments. The furnace temperature was monitored by a B-type thermocouple placed next to the samples with an overall temperature accuracy estimated to be ±2°C. The platinum crucibles, holding specific oxide mixtures, were suspended by a platinum wire (0.5-mm-diameter) in the hot zone of the furnace. All samples were pre-melted again before equilibration at temperature 1650°C for 3 hours, after which the temperature was cooled down to the equilibration temperature (1300°C or 1400°C), an equilibration time of 24 hours based on the results from preliminary experiments and experience reported by previous authors^31,32,33) was employed to ensure equilibrium had been achieved. Ar (99.99 pct pure) gas was passed through the furnace during the whole experiment to avoid potential contamination sources.

After the equilibrium, the base of the furnace was removed prior to rapid quenching of the sample directly into ice water just below the furnace. Samples were then dried and mounted in epoxy resin and polished for analysis. XRD, SEM and Energy Dispersive X-ray spectroscope (EDX, BRUKER) were used to identify the co-existing phase and analyze the composition of each sample. The EDX measurement conditions were as follows: a beam current of 15 nA, an accelerating voltage of electron beam was 20 kV. CaSiO₃, MgO, Al₂O₃ and TiO₂ were respectively used as standard to analyze CaO, SiO₂, MgO, Al₂O₃ and TiO₂. The resolution of the detector was 128 eV. The ZAF correction method was used to calculate the composition of CaO, SiO₂, MgO, Al₂O₃ and TiO₂. For accuracy, six analysis points were randomly chosen for each phase in the samples.

3. Results and Discussion

3.1. The Equilibria Results

With the purpose of clarifing the equilibrium phases change with compositions in CaO-SiO₂-5wt.%MgO-20wt.%Al₂O₃-TiO₂ system, the high temperature equilibrium experiments have been conducted at 1300°C and 1400°C. The microstructure and XRD pattern of the equilibration phases at 1300°C and 1400°C are shown in Figs. 2, 3, 4, 5, and the compositions of the equilibration phases at 1300°C and 1400°C are shown in Tables 2 and 3, respectively. The composition of the quenched phase are the average values calculated from six different analysis points in the samples. In the following sections, L stands for liquid, (C₂MS₂, C₂AS)_ss stands for melilite solid solution, (MT₂, AT)_ss stands for pseudobrookite solid solution.

Fig. 2.

Microstructure of equilibrium phases for CaO-SiO₂-5wt.%MgO-20wt.%Al₂O₃-TiO₂ system at 1300°C. (a) K2 (b) K6 (c) K5 (d) K8 (e) K12 (f) K15.

Fig. 3.

XRD pattern for sample K2, K5, K6, K8, K12 and K15 at 1300°C.

Fig. 4.

Microstructure of equilibrium phases for CaO-SiO₂-5wt.%MgO-20wt.%Al₂O₃-TiO₂ system at 1400°C. (a) K3 (b) K10 (c) K8 (d) K12 (e) K15 (f) K17.

Fig. 5.

XRD pattern for sample K3, K8, K10, K12, K 15 and K17 at 1400°C.

Table 2. Equilibrium phase compositions for CaO-SiO₂-5wt.%MgO-20wt.%Al₂O₃-TiO₂ system at 1300°C.

Sample No.	Contrast in SEM microphotograph	Phase	Compositions, wt.%
			TiO2	SiO2	CaO	MgO	Al2O3	Total
K1	gray	L	6.03	40.88	31.14	4.89	17.06	100.00
K2	dark gray	L	6.46	37.91	34.64	4.04	16.94	100.00
	light gray	melilite solid solution		26.05	35.16		38.79	100.00
K3	light gray	L	6.18	36.81	32.01	6.46	18.54	100.00
	dark gray	melilite solid solution		29.36	36.63	5.07	28.94	100.00
K5	light gray	L	3.92	34.81	36.27	4.08	20.92	100.00
	dark gray	melilite solid solution		31.12	34.67	4.44	29.78	100.00
	white dentric	perovskite	58.74		41.26			100.00
K6	gray	L	8.12	35.09	32.90	4.68	19.21	100.00
	white dentric	perovskite	58.87		41.13			100.00
K7	light gray	L	8.80	36.12	30.16	6.60	18.32	100.00
	dark gray	melilite solid solution		28.59	40.99	3.66	26.76	100.00
	white dentric	perovskite	58.29		41.71			100.00
K8	gray	melilite solid solution		28.57	42.39	2.55	26.49	100.00
	white dentric	perovskite	57.96		40.38		1.66	100.00
K9	light gray	L	11.19	38.28	26.02	4.47	20.04	100.00
	dark gray	Al–Ti diopside	15.49	31.32	24.66	6.35	22.18	100.00
	white dentric	perovskite	58.99		41.01			100.00
K12	light gray	L	9.30	38.16	26.94	5.39	20.21	100.00
	dark gray	Al–Ti diopside	17.97	28.30	25.50	5.51	22.72	100.00
	white dentric	perovskite	58.93		41.07			100.00
K15	light gray	L	18.51	34.26	23.36	4.08	19.79	100.00
	dark gray	Al–Ti diopside	18.57	27.69	24.68	5.56	23.50	100.00
	white dentric	perovskite	58.75		41.25			100.00
	gray strip	(MT2, AT)ss	87.65			6.62	5.73	100.00
K17	light gray	L	10.38	38.52	27.66	3.98	19.46	100.00
	dark gray	Al–Ti diopside	15.48	30.98	25.32	5.98	22.23	100.00
	white dentric	perovskite	58.28		41.72			100.00

Table 3. Equilibrium phase compositions for CaO-SiO₂-5wt.%MgO-20wt.%Al₂O₃-TiO₂ system at 1400°C.

Sample No.	Contrast in SEM microphotograph	phase	Compositions, wt.%
			TiO2	SiO2	CaO	MgO	Al2O3	sum
K3	light gray	L	5.43	30.99	38.58	4.60	20.40	100.00
	dark gray	melilite solid solution		27.89	41.25	3.01	27.85	100.00
K4	light gray	L	7.45	29.49	38.06	5.87	19.13	100.00
	dark gray	melilite solid solution		28.52	42.15	2.96	26.37	100.00
	white	perovskite	58.78		41.22			100.00
K6	gray	L	9.52	31.25	34.63	5.23	19.37	100.00
K8	light gray	L	8.94	29.81	36.25	6.37	18.63	100.00
	dark gray	melilite solid solution		27.60	41.00	2.30	29.10	100.00
	white	perovskite	59.00		41.00			100.00
K10	gray	L	12.79	28.75	33.63	5.04	19.80	100.00
	white	perovskite	58.81		41.19			100.00
K11	light gray	L	12.67	28.88	34.91	5.19	18.35	100.00
	dark gray	melilite solid solution		26.27	42.65	1.47	29.62	100.00
	white	perovskite	58.85		41.15			100.00
K12	light gray	L	18.95	28.93	28.61	4.64	18.87	100.00
	dark gray	Al–Ti diopside	21.03	28.35	25.60	5.53	19.49	100.00
K13	gray	L	15.62	27.49	31.14	5.15	20.61	100.00
	white	perovskite	58.87		41.13			100.00
K14	gray	L	13.66	27.23	31.98	5.47	21.66	100.00
	white	perovskite	58.91		41.09			100.00
K15	light gray	L	22.44	29.76	23.26	4.83	19.71	100.00
	dark gray	Al–Ti diopside	22.33	28.66	24.33	5.45	19.24	100.00
	white strip	(MT2, AT)ss	92.87			2.87	4.26	100.00
K16	light gray	L	22.96	27.18	25.31	4.48	20.07	100.00
	dark gray	Al–Ti diopside	23.21	26.20	23.62	5.87	21.10	100.00
K17	light gray	L	20.00	26.68	26.25	5.62	21.45	100.00
	dark gray	Al–Ti diopside	20.95	26.44	31.76	3.59	17.26	100.00
	white	perovskite	58.89		41.11			100.00

3.1.1. Phase Relations at 1300°C

In Fig. 2(a), two phases were detected by SEM for sample K2, based on the XRD result of K2 in Fig. 3 and the phase compositions in Table 2, it is easy to confirm that the light gray phase is the melilite solid solution (C₂MS₂, C₂AS)_ss phase, and the dark gray phase is the coexisting liquid phase. Figure 2(b) gives a sample (K6) example for liquid-perovskite (CaO·TiO₂) coexisting equilibrium at 1300°C, similarly, the white dendritic phase is perovskite (CaO·TiO₂) and the gray phase is the quenched glass phase based on the combination of the XRD result in Fig. 3 and the compositions in Table 2.

As for sample K5, both the melilite solid solution (C₂MS₂, C₂AS)_ss phase and perovskite (CaO·TiO₂) phase are found to be coexisting with liquid, as shown in Fig. 2(c), the white dendritic phase is perovskite (CaO·TiO₂) phase, the dark gray phase is melilite solid solution (C₂MS₂, C₂AS)_ss phase and the light gray phase is liquid phase. The equilibrium result for sample K7 is the same with K5. When the basicity is further increased to sample K8, the result is somehow different, from the XRD diffraction peak in Fig. 3 and the compositions in Table 2, it can be confirmed that in Fig. 2(d) the white phase is perovskite (CaO·TiO₂) phase and the matrix gray phase is the melilite solid solution (C₂MS₂, C₂AS)_ss phase, that is to say, sample K8 present as solid phase coexistence and there is no liquid existing at 1300°C.

Figure 2(e) gives a example (K12) of liquid coexist with perovskite (CaO·TiO₂) phase and Al–Ti diopside phase, by the same method, the white dendritic phase is confirmed as perovskite (CaO·TiO₂) phase, light gray phase is liquid phase and the dark gray phase is Al–Ti diopside phase. The Al–Ti diopside phase is one of the most common phases presented in Ti-bearing slag system and is characterized by the high solubility of Al₂O₃ and TiO₂.^{34,35,36,37,38,39,40)} In Fig. 2(f), four phases equilibria are detected for sample K15 at 1300°C, except for the equilibrium phases mentioned for sample K12, the pseudobrookite solid solution phase with the fomula of (MT₂, AT)_ss is found, as marked in Fig. 2(f). MgO·TiO₂ and Al₂O₃·TiO₂ are both with the pseudobrookite structure, and (MT₂, AT)_ss phase is also a common phase frequently appeared in Ti-bearing slag system.^41,42,43)

3.1.2. Phase Relations at 1400°C

The equilibria phases presented at 1400°C are not that complicated compared with the results at 1300°C, and the XRD patterns could be found in Fig. 5. In Figs. 4(a) and 4(b), liquid coexists with melilite solid solution (C₂MS₂, C₂AS)_ss phase and liquid coexists with perovskite (CaO·TiO₂) phase is exhibited by sample K3 and K10 respectively. In Fig. 4(c), three phases coexisting of liquid with melilite solid solution (C₂MS₂, C₂AS)_ss phase and perovskite (CaO·TiO₂) phase is shown for sample K8, compared with the equilibria phases at 1300°C, the liquids phase appears at 1400°C.

In Fig. 4(d), liquid phase coexists with Al–Ti diopside phase is shown by sample K12, the light gray phase is liquid phase, while the dark gray phase is Al–Ti diopside phase, which looks like leaf shape with magnified image. As for sample K15, liquid phase coexists with Al–Ti diopside phase and (MT₂, AT)_ss phase is observed in Fig. 4(e), the white phase is (MT₂, AT)_ss phase and the dark gray Al–Ti diopside phase is uniformly distributed in the light gray liquid phase. In the case of sample K17, liquid phase is found to coexist with perovskite (CaO·TiO₂) phase and Al–Ti diopside phase, as marked in Fig. 4(f).

3.2. Construction of the Liquidus Lines

Based on the experimental results mentioned above, the 1300°C and 1400°C liquidus line are constructed for the CaO-SiO₂-5wt.%MgO-20wt.%Al₂O₃-TiO₂ system in Fig. 6. As can be seen, the liquidus temperature increases with the increase of basicity and TiO₂ content. In order to make a convenient discussion and comparison, it is more suitable to plot the composition of the equilibrium solid phases (such as the Al–Ti diopside phase) on the pseudo-ternary phase diagram in Fig. 6. The compositions of the equilibrium solid phases are the average value of the detected compositions of the equilibrium solid phases in Tables 2 and 3. When plotting the compositions, a normalization according to different authors^18,45,46) has been done based on the following steps: the Al₂O₃ and MgO contents were adjusted to 20 wt.% and 5 wt.% respectively, whereas the compositions of other three components were normalized in proportion to their original fractions with their sum being 75 wt.%, after the normalization, the composition of the equilibrium solid phases could be plotted on the ternary system.

Fig. 6.

1300°C and 1400°C liquidus lines for CaO-SiO₂-5wt.%MgO-20wt.%Al₂O₃-TiO₂ system, wt.%.

In Fig. 6, the liquidus lines from Sun et al.²⁵⁾ are plotted as short dash lines for comparison. As can be seen, both the 1300°C liquidus line from present work and Sun et al.²⁵⁾ have the same variation trend with compositions. As for 1400°C isotherm, the liquidus line changes with composition below 15 wt.% TiO₂ from both have the same variation trend, however, the 1400°C isotherm from present work shifts to a little higher solubility of SiO₂ when the TiO₂ content is higher than 15 wt.%. The deviation may be due to the different method employed in the experiments, Sun et al.²⁵⁾ established the phase diagram in dynamic conditions which is different from the static methods adopted in present work, otherwise, the normalization of present data can also cause minor shift. In Fig. 6, the 1300°C and 1400°C isotherms calculated by factsage software⁴⁴⁾ are also added for comparison. It can be seen from Fig. 6 that the calculated 1300°C isotherm has a good accordance with both the 1300°C isotherm from present work and Sun et al.,²⁵⁾ while the calculated 1400°C isotherm has a higher solubility of SiO₂ when the TiO₂ content is smaller than 19 wt.%, even though experimental results and calculated data have the same variation trend with compositions. The difference may arise from the database, as illustrated in the Database Documentation, in the presence of Ti, the database has been developed only for reducing conditions, and the liquid phase is generally well modeled only for binary systems, but for ternary and higher-order systems is only estimated from the model due to lack of any experimental data.

3.3. Phase Diagram of CaO-SiO₂-5wt.%MgO-20wt.%Al₂O₃-TiO₂ System

Coupling the present results with the isothermal information from Sun et al.,²⁵⁾ the phase diagram for the CaO-SiO₂-5wt.%MgO-20wt.%Al₂O₃-TiO₂ system is constructed in Fig. 7. In Fig. 7, the solid lines are the 1300°C to 1500°C liquidus lines, as can be seen, the 1300°C to 1500°C liquidus temperature increases with the increase of basicity and TiO₂ content.

Fig. 7.

Phase diagram for the specific compositions relevent to Ti-bearing slag system, wt.%.

From the equilibrium phases detected at 1300°C and 1400°C, the change of the primary crystal phases could be concluded as follows. In the low TiO₂ content, the primary phase is melilite solid solution (C₂MS₂, C₂AS)_ss phase, while it changed as perovskite (CaO·TiO₂) phase with the TiO₂ content increased. Meanwhile, the dominated phase become as Al–Ti diopside phase in the lower basicity area. The information of isotherms and primary crystal field from present work is important for related processes, take the SCPS (the Selective Crystallization and Phase Separation) method as an example. As mentioned above, the key of SCPS method is to enrich the Ti element into a certain phase, i.e. perovskite, rutile or anosovite. Based on the phase diagram information in Fig. 7, if perovskite is chosen as the Ti-enriched phase, the composition of Ti-bearing slag can be adjusted to the primary crystal phase field by decreasing the TiO₂ content, and decrease the temperature from 1500°C to 1350°C to increase the proportion of perovskite.

4. Conclusions

In order to gain more thermodynamic information for the development of Ti-bearing slag system, the phase equilibrium relationships were investigated for the CaO-SiO₂-5wt.%MgO-20wt.%Al₂O₃-TiO₂ phase diagram system. The equilibrium phases were experimentally determined at 1300°C and 1400°C using the high temperature equilibrium technique followed by X-Ray Fluorescence (XRF), X-Ray Diffraction (XRD), Scanning Electron Microscope (SEM) and Energy Dispersive X-ray spectroscope (EDX) analysis. The liquid phase (L), melilite solid solution (C₂MS₂, C₂AS)_ss phase, perovskite (CaO·TiO₂) phase, Al–Ti diopside phase and (MT₂, AT)_ss phase were found. Based on the experimental results from present work and the information from literatures, the 1300°C to 1500°C liquidus lines and phase diagram were constructed for the specified region of the CaO-SiO₂-5wt.%MgO-20wt.%Al₂O₃-TiO₂ system. The phase equilibrium data is important for the comprehensive utilization of Ti-bearing slag system.

Acknowledgement

This work was supported by National Key R&D Program of China [Nos. 2017YFC0805105], the National Natural Science Foundation of China [Nos. 51374059, 51104039 and 51304042], Scientific Research Fund of Liaoning Provincial Education Department [No. L2013114], Programs of Liaoning province for Science and Technology Development [No. 2012221013], and the Fundamental Research Funds for the Central Universities of China [N130602002].

References

• 1)   W.  Zhang,  L.  Zhang and  J. h.  Zhang: Ind. Eng. Chem. Res., 51 (2012), 12294.
• 2)   X. H.  Liu and  Z. T.  Sui: China Nonferr. Met., 12 (2002), 1281.
• 3)   M. H.  Wang,  X. H.  Du and  Z. T.  Sui: Multipurp. Util. Miner. Resour., (2000), No. 4, 5.
• 4)   I.  Barin and  W.  Schuler: Metall. Trans. B, 11 (1980), 199.
• 5)   P.  Yi: Titan. Ind. Prog., 3 (2005), 44.
• 6)   Y. M.  Wang,  R. F.  Liu,  R. H.  Zhou,  B. S.  Shao,  Q. S.  Wei,  Z. F.  Yuan and  C.  Xu: Comput. Appl. Chem., 23 (2006), 263.
• 7)   X. D.  Wang,  Y. W.  Mao,  X. Y.  Liu and  Y. K.  Zhu: J. Iron Steel Res., 3 (1990), 1.
• 8)   Y. Q.  Sun,  J.  Li,  X. D.  Wang and  Z. T.  Zhang: Metall. Mater. Trans. B, 45 (2014), 1446.
• 9)   M. Y.  Wang,  L. N.  Zhang,  L.  Zhang,  Z. T.  Sui and  G.-F.  Tu: Trans. Nonferr. Met. Soc. China, 16 (2006), 421.
• 10)   J.  Li,  Z.  Zhang,  L.  Liu,  W.  Wang and  X.  Wang: ISIJ Int., 53 (2013), 1696.
• 11)   Y.  Sun,  Z.  Li,  L.  Liu,  X. D.  Wang and  Z. T.  Zhang: ISIJ Int., 55 (2015), 158.
• 12)   X. D.  Wang,  Y. W.  Mao,  D. S.  Xie and  Y. K.  Zhu: J. East China Inst. Metall., 10 (1993), 16.
• 13)   R. C.  DeVries,  R.  Roy and  E. F.  Osborn: J. Am. Ceram. Soc., 38 (1955), 158.
• 14)   A.  Ohno and  H. U.  Ross: Can. Metall. Q., 2 (1963), 243.
• 15)   H. A.  Fine and  S.  Arac: Ironmaking Steelmaking, 7 (1980), 160.
• 16)   E. F.  Osborn and  K. H.  Gee: Bull. Earth Miner. Sci. Exp. Stn., 85 (1969), 57.
• 17)  Slag Atlas, 2nd ed., ed. by VDEh, Verlag Stahleisen, Düsseldorf, (1995), 39.
• 18)   Z.  Wang,  H.  Sun and  L.  Zhang: J. Alloy. Compd., 671 (2016), 137.
• 19)   B. J.  Zhao,  J.  Eugene and  C. H.  Hayes: 2009 Chniese Society of Metals Annual Meeting Proc., Metallurgical Industry Press, Beijing, (2009), 438.
• 20)   H. M.  Liu,  H.  Du and  Z. P.  Yang: Acta Metall. Sin., 28 (1992), 47.
• 21)   S. M.  Jung and  R. J.  Fruehan: ISIJ Int., 41 (2001), 1447.
• 22)   Y.  Morizane,  B.  Ozturk and  R. J.  Fruehan: Metall. Mater. Trans. B, 30 (1999), 29.
• 23)   J. J.  Shi,  L. F.  Sun,  J. Y.  Qiu and  M. F.  Jiang: Metall. Mater. Trans. B, 47 (2016), 425.
• 24)   J. J.  Shi,  L. F.  Sun,  J. Y.  Qiu and  M. F.  Jiang: J. Cent. South Univ. (Sci. Technol.), 47 (2016), 3309.
• 25)   L. F.  Sun,  J. J.  Shi and  J. Y.  Qiu: J. Cent. South Univ., 24 (2017), 48.
• 26)   J. J.  Shi,  L. F.  Sun and  J. Y.  Qiu: ISIJ Int., 56 (2016), 1124.
• 27)   J. J.  Shi,  L. F.  Sun and  J. Y.  Qiu: J Alloy. Compd., 699 (2017), 193.
• 28)   Y.  Kashiwaya,  C. E.  Cicutti,  A. W.  Cramb and  K.  Ishii: ISIJ Int., 38 (1998), 348.
• 29)   N.  Nityanand and  H.  Alan Fine: Metall. Trans. B, 14 (1983), 685.
• 30)   S.  Ueda,  K.  Takemoto,  T.  Ikeda and  M.  Maeda: ISIJ Int., 40 (2000), S92.
• 31)   H. M.  Henao,  K.  Florian and  K.  Itagaki: Mater. Trans., 46 (2005), 812.
• 32)   Y.-B.  Kang and  H.-G.  Lee: ISIJ Int., 45 (2005), 1552.
• 33)   H. M.  Henao,  C.  Pizarro,  J.  Font,  A.  Moyano and  P. C.  Hayes: Metall. Mater. Trans. B, 41 (2010), 1186.
• 34)   A. E.  Rubin: Meteor. Planet. Sci., 42 (2007), 1711.
• 35)   A. N.  Krot,  E. R. D.  Scott and  M. E.  Zolensky: Meteoritics, 30 (1995), 748.
• 36)   N.  Morimoto: Mineral. Petrol., 39 (1988), 55.
• 37)   A. K.  Gupta,  K.  Onuma and  K.  Yagi: Contrib. Mineral. Petrol., 41 (1973), 333.
• 38)   H. X.  Fang,  H. Y.  Li and  T.  Zhang: ISIJ Int., 55 (2015), 200.
• 39)   M. Y.  Wang,  L.  Zhang and  Z. T.  Sui: Ironmaking Steelmaking, 36 (2009), 388.
• 40)   L.  Zhang,  W.  Zhang and  J.  Zhang: Int. J. Miner. Metall. Mater., 23 (2016), 1029.
• 41)   B.  Hong and  W.  Fu: Multipurp. Util. Miner. Resour., (2012), No. 3, 55.
• 42)   X.  Goso,  J.  Nell and  J.  Petersen: Advances in Molten Slags, Fluxes and Salts, Proc. MOLTEN 2016, Springer, New York, (2016), 105.
• 43)   X. D.  Wang,  Y. W.  Mao and  X. Y.  Liu: J. Iron Steel Res., 3 (1990), 1.
• 44)   C. W.  Bale,  E.  Bélisle and  P.  Chartrand: Calphad, 54 (2016), 35.
• 45)   G.  Jimmy,  B. J.  Yan and  S. C.  Du: Metall. Mater. Trans. B, 42 (2011), 1008.
• 46)   G.  Jimmy,  Y. L.  Wang and  S. C.  Du: Calphad, 35 (2011), 249.