Generation of Pyroxene-Based Porous Ceramics from Steel Refining Slag

Abstract

In this paper, porous ceramics were prepared by controlling the composition and sintering temperature of raw materials based on slag from steel refining operations and without the need for a pore-forming agent. Solid-state processing included a phase transformation from low-density raw materials (2.10–2.90 g/cm³) to high-density pyroxene-based ceramics (3.22–3.88 g/cm³). Phase transformations and densification of porous ceramics were investigated by means of mercury intrusion, XRF, SEM and EDS. Results showed that a porous ceramic with optimum properties produced from raw material containing 35 wt.% steel refining slag, exhibited a high water absorption capacity of 31.62%, flexural strength of 18.26 MPa, porosity of 44.5% and an average pore diameter of 1.27 µm. Raw materials with high fineness or content near the pyroxene composition promoted formation of pyroxene and accordingly more volumetric shrinkage which enhanced both bending strength and porosity of the porous ceramics.

1. Introduction

About one million tonnes of slag are generated each year in the steelmaking process for a large iron and steel corporation with an annual output of 10 million tonnes crude steel, of which approximately 90% is generated under an oxidizing atmosphere. A number of studies have been conducted to explore methods for using the slags produced under oxidizing conditions. Examples include: recovery of iron from steelmaking slags,¹⁾ use as raw material for glass and ceramic industries,^2,3) road and hydraulic construction,⁴⁾ fertilizers and soil improvement.⁵⁾

The slag, as a by-product generated from refining processes conducted under reducing conditions, which is defined as steel refining slag in this work, contains more CaO (45 wt.%–55 wt.%) than normal steelmaking slag (about 40 wt.%CaO and a certain amount FeO), and is hard to reuse in the cement or concrete industries due to its instability derived from α-Ca₂SiO₄, free CaO and free MgO. The objective of this research is to find additional applications for steel refining slag with high CaO content.

Porous ceramics are a new type of functional material having high porosity, high specific surface area, high temperature resistance, corrosion resistance and good stability, and widely used as filter materials, absorber materials, thermal insulation materials and catalyst carriers.^6,7) The raw materials for preparing porous ceramics vary according to their application, and include SiC, Si₃N₄ and Al₂O₃,^{8,9,10,11,12)} as well as metallurgical slag, waste glass and construction waste.^{13,14,15,16,17)}

There are three methods generally used to produce porous ceramics. The first method consists of the addition of a pore-forming agent, most commonly, a volatile or flammable material which volatilizes from the matrix during high-temperature sintering and thereby pores are formed.¹⁸⁾ However, the ceramics prepared by this method generally have a porosity of less than 50% and poor pore distribution. The second method involves the addition to the raw material of an organic or inorganic substance which acts as a foaming agent, and volatilizes to produce a foam under high-temperature conditions.^2,19) The third method is based on a gelation process in which colloidal ions are interconnected to form a spatial network structure,²⁰⁾ the pores in which are filled with solution. This solution evaporates during the sintering process, leaving many nano-sized or micro-sized voids in the ceramic material. However, this method is complicated, with high cost, and therefore is only used for some special applications.

Traditional ceramics consist of clay, feldspar and quartz and contain a final new phase of mullite. Ceramics which use shale, fly ash or coal gangue, mainly have crystals of quartz and mullite.²¹⁾ Cheng et al.²²⁾ used clay, quartz and alumina as the main raw materials and converted them into ceramics with the main phase being anorthite. Other researchers^23,24,25) have conducted experiments on producing ceramics from steel slag. Zhao et al.²⁶⁾ used steel slag, talc, pyrophyllite and clay as the main raw materials to prepare ceramics in which the main crystals were pyroxene. During the sintering process, meta-stable phases in slags, such as Ca₂SiO₄, free CaO and free MgO can transform into stable phases such as CaMgSi₂O₆ and CaAl₂Si₂O₈.²⁷⁾

Pyroxene minerals including diopside, augite and hedenbergite constitute an ideal high-density phase and have the general formula of M₂M₁T₂O₆ with a wide isomorphism involving Ca²⁺, Na⁺, Mn²⁺ and Mg²⁺ at position M₂, Fe²⁺, Mg²⁺, Mn²⁺, Ni²⁺, Li⁺, Al³⁺, Fe³⁺, Cr³⁺, and Ti³⁺ at position M₁, and Si⁴⁺, Al³⁺ at position T. Thus the main raw materials for the preparation of ceramics can be derived from solid waste or inferior minerals containing CaO, MgO or Fe₂O₃.²⁸⁾

Crystal phase transition is accompanied by a volume change when there is a difference in crystal density before and after the reaction. For example, quartz, feldspar and enstatite can react to form pyroxene. The densities of quartz, akermanite and pyroxene are 2.22–2.65 g/cm³, 2.90–3.10 g/cm³ and 3.22–3.88 g/cm³, respectively. When 1 mol of pyroxene is produced, the volume of the crystal phase will shrink by 19.17%. If a porous ceramic material is prepared using such a volume change, the pore-forming and foaming agents are not required.

Studies^23,28) have shown that the production of pyroxene-based ceramics containing steel slag had a larger shrinkage of approximately 12% compared to 7–10% associated with ceramics produced by traditional methods. This suggests that the preparation of porous ceramics using materials which undergo an appropriate volume change during the sintering process would eliminate the need for pore-making agents.

From previous studies of different ceramics, pyroxene-based ceramic was selected for this work because of its high density. Since steel refining slag has a high CaO content and a low Fe₂O₃ content, it could be an appropriate raw material to produce pyroxene-based ceramics. With this objective, different batches of pyroxene ceramics were prepared from raw materials with different fineness and at different sintering temperatures and the effects of crystals on the porous structure as well as the formation mechanism of pores in the ceramics were investigated.

2. Experimental Aspects

2.1. Raw Materials

Raw materials including steel refining slag (Jin Hui Steelmaking Plant, Henan Province, China), clay and dolomite (Shandong Yi Ke Co. Ltd., Shandong Province, China), and quartz and pyrophyllite (Foshan Oceano Ceramics Co. Ltd., Guangdong Province, China) were analyzed using X-ray fluorescence spectrometry.

The chemical compositions of the various materials are presented in Table 1. The MgO and CaO were mainly derived from the slag and the calcined dolomite. The compositions of both the mixture and the improved sample lie within the SiO₂–Al₂O₃–CaO–MgO system, however they differ in CaO+MgO and SiO₂+Al₂O₃ contents. The slag, clay, quartz, pyrophyllite and dolomite have densities of 2.70–2.90 g/cm³, 2.10–2.20 g/cm³, 2.22–2.65 g/cm³, 2.80–2.90 g/cm³ and 2.80–2.90 g/cm³ respectively.

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

Materials	CaO	SiO2	MgO	Al2O3	Fe2O3	CaF2	MnO	K2O	Na2O	Total
Steel refining slag	45.66	28.25	8.64	2.19	0.71	5.8	6.68	0	0	97.93
Clay	2.48	69.03	3.21	17.38	1.86	0	0	4.15	0.75	98.86
Quartz	0	95.21	0.16	2.83	0.42	0	0	1.03	0	99.65
Pyrophyllite	0.02	83.45	0.04	13.57	0.33	0	0	2.09	0.21	99.71
Calcined Dolomite	51.3	7.1	39.5	1.9	0	0	0	0	0	99.8
Mixture sample	17.07	61.52	3.69	9.54	0.72	2.12	2.44	1.61	0.21	98.92
Improved sample	26.50	51.53	11.30	3.86	0.58	2.03	2.34	0.78	0.08	99.00

2.2. Preparation of Porous Ceramics

The samples were prepared from a mixture with 36.5 wt% steel refining slag, 16.0 wt% clay, 43.0 wt% quartz, and 4.5 wt% pyrophyllite. The samples were mixed with water using the weight ratio of 1:1 and milled for 20 min, 25 min, 30 min, and 50 min using a planetary ball mill. The four samples from course to fine particle size were identified as #1, #2, #3 and #4.

An improved sample after analysis of the above four samples was designed and prepared from a mixture with 35 wt% steel refining slag, 10 wt% clay, 35 wt% quartz, and 20 wt% calcined dolomite. This sample was processed in the same way as the above samples and with a milling time of 30 min.

The obtained slurries were passed through an 80-mesh sieve and dried. The dried samples were further crushed and screened with a 20-mesh sieve. Batches of 50 mm × 5 mm × 5 mm samples were molded from the sieved powders, dried at 105°C for 12 h., placed in a gradient furnace and heated at a rate of 5°C/min to the desired sintering temperature and held for 1 h.

The initial and sintered #4 samples were respectively heated to the different desired temperatures: 700°C, 800°C, 900°C, 1000°C, 1020°C, 1040°C, 1060°C and 1080°C, and identified as #4-raw, #4-700, #4-800, #4-900, #4-1000, #4-1020, #4-1040, #4-1060 and #4-1080.

2.3. Material Performance Test

Particle sizes of the slurries were determined using a laser particle size distribution analyzer. Bending strength of the fired rectangular samples were evaluated using a ceramic tile flex test. The samples having a section height h (mm), width b (mm) and fulcrum spacing l (mm) were measured using a digital Vernier caliper, and the bending flex f (MPa) calculated using Eq. (1).   

$$f=3Fl/(2b{h}^2)$$(1)

The samples were subjected to a water absorption test using a ceramic water absorption vacuum device. The parameters W₀ (g) and W₁ (g) represent the sample weights before and after water absorption.   

$$W=(W_1-W_0)/W_0$$(2)

The crushed powders were scanned using X-ray diffraction. After coating with Au, the samples were examined by scanning electron microscopy. The porosity and pore size of the samples were determined using a mercury intrusion meter.

3. Results and Discussion

3.1. Properties of Raw Materials

The XRD patterns of the slag and calcined dolomite are shown in Fig. 1. The main mineral phases in the slag are akermanite (Ca₂Mg(Si₂O₇)), merwinite (Ca₃Mg(SiO₄)₂), and calcium silicate (Ca₂SiO₄). The calcined dolomite consisted of periclase (MgO), portlandite (Ca(OH)₂) and lime (CaO).

Fig. 1.

XRD patterns of steel refining slag and calcined dolomite. (Online version in color.)

Samples #1, #2, #3 and #4 had the same compositions as the mixture sample shown in Table 1 but they had different particle size distributions as indicated in Fig. 2. It was found that their median diameters D50 decreased with increasing milling time from 8.262 μm, 6.352 μm, 5.431 μm to 4.248 μm.

Fig. 2.

Particle size distribution of raw materials. (Online version in color.)

3.2. Effect of Particle Size on Densification and Crystallization of Porous Ceramics

The water absorption and flexural strength of the ceramic samples #1, #2, #3 and #4 sintered at different temperatures are shown in Fig. 3. At 700°C, all the samples had almost the same water absorption and flexural strength except that the finer samples #3 and #4 had slightly lower values of water absorption than those of #1 and #2. When the sintering temperature increased from 1020°C to 1060°C, water absorption and flexural strength of the four samples had little change. Moreover, #3 and #4 had similar water absorption that was significantly higher than #1 and #2, while #4 had the highest flexural strength, and #3 had a value close to #2 but obviously higher than #1.

Fig. 3.

Physical properties of ceramics obtained from raw materials with different particle sizes. a) Variation of water absorption capacity with sintering temperature. b) Variation of flexural strength with sintering temperature. (Online version in color.)

After the sintering temperature was increased beyond 1060°C, the water absorption of #3 and #4 was abruptly decreased and lower than that of #1 and #2. Also their flexural strength became much higher than that of #1 and #2. As the densification process entered into a liquid sintering stage, the water absorption significantly decreased and the mechanical properties accordingly improved.

It is interesting that within the temperature range 1020–1060°C, before the liquid sintering stage, samples #3 and #4 had higher water absorption and flexural strength than those of samples #1 and #2. This is unusual since a higher flexural strength is generally associated with lower water absorption.

As shown in Fig. 4, the four samples sintered at 1020°C have similar crystal phases: quartz (SiO₂), anorthite (Ca(Al₂Si₂O₈)), diopside (CaMg(SiO₄)₂) and akermanite (Ca₂Mg(Si₂O₇)), although the relative intensities of the characteristic peaks for diopside and akermanite were quite different. From samples #1 through #4, as the raw materials became finer, the relative intensity of the characteristic peak for diopside increased, but that for akermanite decreased as well as that for quartz. This means that a transformation from akermanite to diopside occurred at this temperature, and the finer particles promoted this transformation.

Fig. 4.

XRD patterns of different samples sintered at 1020°C. (Online version in color.)

A previous study²⁸⁾ focusing on steel slag ceramics showed that akermanite will be converted to pyroxene above 1000°C. In the reaction, akermanite further reacted with silicon-rich and magnesium-rich components at 1020°C and was converted into diopside which is a typical pyroxene containing higher contents of CaO and MgO.

The densities of akermanite (2.90–3.10 g/cm³) and diopside (3.22–3.56 g/cm³) are quite different from each other. As a reaction product, diopside has a 13% higher density than that of akermanite. Due to this difference, the transformation will lead to a shrinkage in volume. As shown in Fig. 3, solid state sintering was dominant at 1020°C. Unlike liquid phase sintering, the reaction area rather than reaction temperature is one of the most important factors for solid state sintering because of the limited diffusion.

Therefore, in the case of samples #3 and #4 with finer particles and more reaction area, more reaction will lead to stronger bonding and higher flexural strength. Figure 4 confirms that more diopside will form with increase in fineness of the raw materials. As a result of the volume shrinkage of crystals associated with the transformation, the increased formation of pores will lead to higher water absorption.

3.3. Effect of Crystallization on Pore Structure of Sample #4

From Fig. 5, it can be observed that the water absorption and flexural strength of sample #4 at 1060°C were 30.2% and 15.97 MPa, respectively. The sample had a narrow pore size distribution (Fig. 6) in the range of 0.6–1.5 μm, with an average pore diameter of 1.02 μm and a porosity of 42.10%. From the SEM image (Fig. 7), it was found that there are a large number of connected pores many of which have diameters of approximately 1 μm or less although several larger pores with a diameter of approximately 4 μm were also observed.

Fig. 5.

Water absorption and flexural strength of Sample #4 at different sintering temperatures. Optimum sintering temperature range is shown by dotted box. (Online version in color.)

Fig. 6.

Pore size distribution of Sample #4 sintered at 1060°C.

Fig. 7.

SEM image of Sample #4 sintered at 1060°C.

The XRD pattern of samples sintered at different temperatures are shown in Fig. 8. The main minerals in sample #4 were quartz, pyrophyllite, merwinite, and akermanite. At 700°C, pyrophyllite will dehydrate to form aluminum silicate. At 800°C, with the decrease of merwinite, akermanite increased accompanied by the formation of calcium silicate, as shown by the arrow in Fig. 8. At 1000°C, merwinite disappeared, while peaks of akermanite reached the highest level. At 900°C, diopside began to form. At 1020°C, the akermanite peaks almost disappeared while those of diopside increased.

Fig. 8.

a) XRD patterns of sample #4 sintered at 700–1020 °C. b) XRD pattern of sample #4 sintered at 1020–1080 °C. The main mineral phases observed in Fig. 8 are as follows. A-Quartz low-SiO₂, B-Pyrophyllite-Al₂Si₄O₁₀(OH₂)₂, C-Akermanite-Ca₂Mg(Si₂O₇), D-Merwinite-Ca₃Mg(SiO₄)₂, E-Aluminum Silicate-Al₂Si₄O₁₀, F-Calcium Silicate-Ca₈Si₅O₁₈, G-Diposide-CaMg(SiO₄)₂, H-Anorthite-Ca(Al₂Si₂O₈), I-Augite-Ca(Mg_0.85Al_0.15)((Si_1.70Al_0.30)O₆). (Online version in color.)

When the sintering temperature increased from 1020°C to 1040°C, the formation of diopside increased significantly. From 1040 to 1060°C augite appeared. Both diopside and augite belong to pyroxene minerals. The appearance of augite with more solid solution of Al³⁺ meant that high sintering temperature is benefit to diffusion of Al³⁺ into diopside and the augite become an aluminum-containing pyroxene. The change from diopside to augite in firing process was also observed by Zhao.²⁸⁾ Since diopside and augite have similar densities, there was no significant shrinkage during this transformation process.

The SEM image of sample #4 sintered at 1060°C is shown in Fig. 9. Selected locations (indicated by areas 1 through 9 in Fig. 9) were analyzed using the EDS technique and the results are listed in Table 2. According to the XRD analysis shown in Fig. 8, it can be deduced that the crystals in Regions 1 and 2 were quartz, while those in Regions 3, 4, 5 and 6 were short columnar crystals of diopside and augite. The flaky crystals in Regions 7, 8, and 9 were anorthite. The particles around pores were mainly short columnar crystals of diopside.

Fig. 9.

SEM image of Sample #4 sintered at 1060°C.

Table 2. Elemental composition (mol) of various points shown in Fig. 9.

Point	Ca	Si	Mg	Al	O
1	0.62	39.02	0.45	0.43	59.49
2	0.40	41.48	0.13	0.26	57.74
3	11.54	15.22	3.90	1.44	67.91
4	13.47	17.24	3.88	1.52	63.90
5	10.85	14.71	4.38	1.17	68.88
6	10.64	14.89	4.74	3.66	66.07
7	8.99	18.81	1.43	9.51	61.26
8	6.43	23.29	0.60	11.86	57.82
9	8.33	17.32	1.33	7.59	65.43

The phase transitions shown in Fig. 8 indicate that the transformation from akermanite to diopside was completed at 1020°C. Referring to the water absorption curve in Fig. 3, it is found that the rate of water absorption did not change when the sintering temperature was increased from 700°C to 1020°C. This suggests that within this temperature range, the formation of higher density phases, such as diopside, accompanied by a decrease in volume would offset the densification due to solid-state sintering. Accordingly, the internal porosity remained unchanged, although the solid-state sintering process increased the flexural strength. The main factors affecting the properties of samples at this stage were the solid-state reaction for both the densification and crystallization processes.

3.4. Effect of Composition on Pore Structure

According to the above analysis, phase transformation of low-density components to high-density phases, such as pyroxene, contributed to the generation of pores. In order to validate this mechanism for the intrinsic formation of pores, a new improved composition with more formation of pyroxene was designed. Based upon the position of pyroxene in the SiO₂–Al₂O₃–CaO–MgO phase diagram,²⁹⁾ the composition was improved by increasing the content of MgO and CaO, while at the same time, decreasing the content of Al₂O₃ and SiO₂, as indicated in Table 1.

The water absorption and flexural strength curves of the improved samples are shown in Fig. 10. Samples sintered within the temperature range 1200–1250°C entered the transitional region from solid-state sintering to liquid phase sintering. The sample sintered at 1220°C had a relatively high flexural strength of 18.26 MPa and a water absorption of 31.62%. Further analysis indicated the sample had a porosity of 44.5% with an average pore diameter of 1.27 μm, and a pore size range of 0.5–1.8 μm.

Fig. 10.

Water absorption and flexural strength curves of improved sample sintered at different temperatures. Optimum sintering temperature range is shown by dotted box. (Online version in color.)

As shown in Table 3, compared to Sample #4, the improved sample had better water absorption, flexural strength and porosity. The XRD pattern of the improved sample sintered at 1220°C is shown in Fig. 11. In contrast to the phases shown for sample #4 in Fig. 8, pyroxene (augite) was the predominant phase in the improved sample. Formation of more pyroxene in the improved sample would result in greater volume shrinkage, which in turn would contribute to higher porosity and water absorption in comparison to sample #4.

Table 3. Comparison of the performance of sample obtained by sintering the improved sample at 1220°C and the original sample #4 at 1060°C.

Samples	Water absorption (%)	Flexural strength (MPa)	Porosity (%)	Average pore diameter (μm)
Mixture sample (#4)	30.2	15.97	42.10	1.02
Improved sample	31.62	18.26	44.50	1.27

Fig. 11.

XRD pattern of the improved sample sintered at 1220°C.

It is evident that there are two requirements for intrinsic formation of pores. One is the transformation of low-density components to high-density phases, such as pyroxene. The other is the limited sintering ability at the beginning of the transitional region from solid state to liquid phase sintering. During the solid-state sintering process, the flexural strength was inadequate. When sintering of samples entered the liquid-state stage at about 1200°C, the diffusion ability of cations significantly increased and this generated a significant increase in strength at the expense of porosity.

Therefore, the optimum sintering condition was the transitional region from solid-state sintering to liquid phase sintering. The optimum temperature range is shown by the red dotted box in Fig. 10. At the sintering temperature of 1220°C, the ceramic produced with the improved composition, exhibited both high porosity and high flexural strength.

4. Conclusions

In this work, slag from steel refining was used to produce a novel porous ceramic without the addition of pore-making agents. The conclusions are as follows:

(1) The porous ceramic with 35 wt% slag exhibited good properties, having water absorption of 31.62%, flexural strength of 18.26 MPa, porosity of 44.5% and an average pore diameter of 1.27 μm.

(2) The porous ceramics made from slag were sintered by solid-state processing that involved phase transformation from low-density raw material (2.10–2.90 g/cm³) to high-density pyroxene (3.22–3.88 g/cm³). The volume shrinkage associated with the phase transformation was confined to the micro reaction areas due to the limited diffusion of cations in the solid state, thereby resulting in the formation of more pores.

(3) A significant increase in strength at the expense of porosity was obtained by increasing the temperature. The optimum sintering condition occurred at temperatures within the transitional region between solid and liquid state, producing a porous ceramic with a relatively high strength as well as porosity.

(4) Increasing formation of pyroxene correlated with more volumetric shrinkage which resulted in high porosity and water absorption. Raw materials with high fineness and composition close to pyroxene promoted formation of pyroxene which contributed to improved bending strength and porosity of the generated ceramics.

(5) The proposed process for the production of porous ceramic materials based on the use of slag and the intrinsic formation of pores eliminates the need for pore forming agents and is a potential low-cost route for the manufacture of micro-porous ceramics.

Acknowledgements

This work has been financially supported by National Natural Science Foundation of China (No. U1960201) and Major Science and Technology Innovation Engineering Projects of Shandong Province (No. 2019JZZY010404).

References

• 1)   A.  Semykina,  V.  Shatokha and  S.  Seetharaman: Ironmaking Steelmaking, 37 (2010), 536.
• 2)   W. Q.  Fang,  L. J.  Hou and  Y.  Li: ISIJ Int., 61 (2021), 1043. https://doi.org/10.2355/isijinternational.ISIJINT-2020-271
• 3)   Y.  Li,  Y. P.  Ren,  D. J.  Pei,  H. J.  Sheng,  Y. D.  Yi and  D. Q.  Cang: ISIJ Int., 59 (2019), 1723.
• 4)   V. A.  Nunes and  P. H. R.  Borges: Constr. Build. Mater., 281 (2021), 122605. https://doi.org/10.1016/j.conbuildmat.2021.122605
• 5)   C. M.  Du,  X.  Gao,  S.  Ueda and  S.  Kitamura: ISIJ Int., 58 (2018), 833.
• 6)   A. R.  Studart,  U. T.  Gonzenbach,  E.  Tervoort and  L. J.  Gauckler: J. Am. Ceram. Soc., 89 (2006), 1771.
• 7)   J. M.  Tulliani,  M.  Lombardi,  P.  Palmero,  M.  Fornabaio and  L. J.  Gibson: J. Eur. Ceram. Soc., 33 (2013), 1567.
• 8)   J.  Liu,  X. B.  Zhang,  H. J.  Wu,  J. X.  Hu,  Z. L.  Wan and  Y.  Feng: Refract. Ind. Ceram., 55 (2015), 545.
• 9)   N.  Li,  X. Y.  Zhang,  Y. N.  Qu,  J.  Xu,  N.  Ma,  K.  Gan,  W. L.  Huo and  J. L.  Yang: J. Eur. Ceram. Soc., 36 (2016), 2807.
• 10)   N.  Dong,  X. N.  Sun,  Y. Z.  Ma,  X. L.  Li and  X. W.  Yin: Ceram. Int., 42 (2016), 11982.
• 11)   N.  Ma,  L. J.  Du,  W. T.  Liu,  X. Y.  Zhang,  W. L.  Huo,  Y. N.  Qu,  K.  Gan,  Y. L.  Wang and  J. L.  Yang: Ceram. Int., 42 (2016), 9145.
• 12)   F.  Wang,  J. W.  Yin,  D. X.  Yao,  Y. F.  Xia,  K. H.  Zuo,  J. Q.  Xu and  Y. P.  Zeng: Mater. Sci. Eng. A, 654 (2016), 292.
• 13)   J. W.  Cao,  J. S.  Lu,  L. X.  Jiang and  Z.  Wang: Ceram. Int., 42 (2016), 10079.
• 14)   C. S.  Zhang,  X.  Wang,  H. J.  Zhu,  Q. S.  Wu,  Z. C.  Hu,  Z. Z.  Feng and  Z.  Jia: Ceram. Int., 46 (2020), 23623.
• 15)   P. R.  Monich,  A. R.  Romero,  D.  Höllen and  E.  Bernardo: J. Clean. Prod., 188 (2018), 871.
• 16)   C. B.  Liu,  L. B.  Liu,  K. F.  Tan,  L. H.  Zhang,  K. J.  Tang and  X. P.  Shi: Ceram. Int., 42 (2016), 734.
• 17)   P.  Sun and  Z. C.  Guo: Trans. Nonferr. Met. Soc. China, 25 (2015), 2230.
• 18)   J. H.  Eom,  Y. W.  Kim,  I. H.  Song and  H. D.  Kim: J. Eur. Ceram. Soc., 28 (2008), 1029.
• 19)   Y.  Zhao,  Z. F.  Zhu,  R. H.  He and  J. F.  Hu: Ceramics, 7 (2008), 27 (in Chinese).
• 20)   T.  Tomita,  S.  Kawasaki and  K.  Okada: J. Porous Mater., 11 (2004), 107.
• 21)   C. Y.  Chen,  G. S.  Lan and  W. H.  Tuan: Ceram. Int., 26 (2000), 715.
• 22)   X. S.  Cheng,  S. J.  Ke,  Q. H.  Wang,  H.  Wang,  A. Z.  Shui and  P. A.  Liu: Ceram. Int., 38 (2012), 3227.
• 23)   Y.  Li,  L. H.  Zhao,  Y. K.  Wang and  D. Q.  Cang: Int. J. Miner. Metall. Mater., 25 (2018), 413.
• 24)   P.  Ter Teo,  A. S.  Anasyida,  C. M.  Kho and  M. S.  Nurulakmal: J. Clean. Prod., 241 (2019), 118144.
• 25)   Y. B.  Zong,  X. D.  Zhang,  E.  Mukiza,  X. X.  Xu and  F.  Li: Appl. Sci., 8 (2018), 1187.
• 26)   L. H.  Zhao,  Y.  Li,  Y. Y.  Zhou and  D. Q.  Cang: Mater. Des., 64 (2014), 608.
• 27)   X. B.  Ai,  Y.  Li,  D. L.  Guo,  L. H.  Zhao and  D. Q.  Cang: J. Cent. South Univ., 46 (2015), 1583 (in Chinese).
• 28)   L. H.  Zhao,  Y.  Li,  L. L.  Zhang and  D. Q.  Cang: ISIJ Int., 57 (2017), 15.
• 29)   V. D.  Eisenhüttenleute: Slag Atlas, Verlag Stahleisen GmbH, Düsseldorf, (1995), 156.