Enhanced Thermal Shock Resistance of Porous Ca₂Mg₂Al₂₈O₄₆ Ceramic Filter via Nano-Sized ZrO₂ Toughening

1. Introduction

The rapid development of high-quality steel products requires efficient and multifunctional filtration materials that possess excellent filtration capacity, thermal stability, and mechanical robustness [1,2,3,4,5]. Various three-dimensional (3D) reticulated ceramics (e.g., SiC, Al₂O₃ and CaO) have been widely applied to remove the non-metallic inclusions (mainly Al₂O₃ inclusions) owing to their intricate pore network and high refractoriness [6,7,8,9,10,11,12]. However, the inadequate filtration capacity, poor thermal shock resistance (TSR) and being prone to hydration limit their broad applications. Recently, Ca₂Mg₂Al₂₈O₄₆ (C₂M₂A₁₄) based ceramics have shown greater potential for filtering Al₂O₃ inclusions because they contain non-free calcium oxide (CaO) that would react with Al₂O₃ inclusions [13,14,15,16]. Nevertheless, the relatively high thermal expansion coefficient (TEC) and low thermal conductivity of C₂M₂A₁₄ deteriorate its TSR. Therefore, the promotion of the TSR of porous C₂M₂A₁₄ ceramics is an urgent task.

Recently, the introduction of the second phase in porous ceramics has been demonstrated to be a versatile approach for enhancing their TSR [9,17,18]. In particular, the incorporation of nano-sized reinforced phase provides a new opportunity to promote their TSR [17]. Among various second phases for improving TSR, zirconia (ZrO₂) has attracted widespread attention due to its enhanced fracture toughness through stress-induced tetragonal-monoclinic transformation [18,19]. For instance, Chen et al [20] prepared porous Al₂O₃-ZrO₂-mullite composites with enhanced TSR through the transformation toughening of ZrO₂ originating from the decomposition of ZrSiO₄. Mao et al [21] optimized the TSR of porous Si₃N₄-based ceramics by adding ZrO₂ powder as the reinforcing phase. However, the direct introduction of these coarse and large-sized ZrO₂ particles would inevitably lead to insufficient phase transformation and uneven distribution of the ZrO₂ in porous ceramic, limiting their further application.

Herein, we demonstrate a robust method for fabricating highly porous C₂M₂A₁₄ ceramics with improved TSR by using ZrO₂ sol as the reinforced composition. The homogeneously distributed nano-sized ZrO₂ particles originating from ZrO₂ sol effectively promote densification and phase transformation, thereby enhancing mechanical properties and TSR of porous ceramics. The as-prepared C₂M₂A₁₄ ceramics demonstrate the integrated properties of relatively high porosity (81.12%), robust cold compressive strength (2.15 MPa), good residual strength ratio (66.4%) and excellent filtration efficiency (68.4%). These results demonstrate that the proposed C₂M₂A₁₄ reticulated ceramics show great potential for application in molten steel purification.

2. Experimental

2.1. Materials and Fabrication Process of Porous C₂M₂A₁₄ Ceramics

The C₂M₂A₁₄ powder (d₅₀ = 20 μm) was synthesized in our previous work [22]. The ZrO₂ sol was purchased from Dezhou Jinghuo Technology Glass Co., Ltd., while polyurethane foam (PU, 15 PPI) was supplied by Wuxi Chenguang Refractory Materials Co., Ltd. Figure 1 shows the XRD patterns and SEM images of ZrO₂ sol. It can be seen that the ZrO₂ sol is amorphous, with its microstructure consisting of agglomerated spherical particles 20~30 nm in size. The solid content of the ZrO₂ sol is 30 wt% and the pH value is 2~4. The carboxymethyl cellulose (CMC), ammonium lignosulfonate (AL) and polycarboxylate (WSM-M) were obtained from Shanghai Aladdin Biochemical Technology Co., Ltd., which were used as the thickener, binder and dispersant, respectively. They are collectively referred to as additives and are added to the ceramic slurry after mixing. In order to enhance the surface roughness of the PU foam and improve the adhesion of the ceramic slurry, the PU foam was pretreated by immersion in a 5 wt% NaOH solution at 25 °C for 24 h, based on relevant literature [23,24]. It was then washed with high-purity deionized water and dried.

Figure 2 schematically depicts the fabrication procedure of the porous C₂M₂A₁₄ ceramics. Firstly, the C₂M₂A₁₄ ceramic powders, CMC, AL and WSM-M powders were mixed with deionized water and underwent high-speed mechanical stirring for 5 min to obtain the homogeneous ceramic slurry. The ZrO₂ sol was added to the slurry, followed by an additional 3 min of high-speed stirring to ensure its uniform dispersion. Unless mentioned, the stirring speed used in this experiment is 2000 r/min. Subsequently, the pretreated PU foam (45 mm × 45 mm × 20 mm) was then immersed in the prepared ceramic slurry and pressed with tweezers to ensure complete impregnation. Excess slurry was removed using a squeeze-roller device. This impregnation-squeezing cycle was repeated three times to obtain the porous C₂M₂A₁₄ ceramics green body. The green body was first dried naturally for 24 h, then dried at 110 °C for another 6 h. Finally, the dried green body was calcined at 1600 °C for 2 h. According to the addition amount of ZrO₂ sol (0, 1, 2, 3 and 4 wt%), the corresponding specimens were labeled as ZS0, ZS1, ZS2, ZS3 and ZS4 (Table 1), respectively.

2.2. Immersion Tests

To evaluate the purification effect of porous C₂M₂A₁₄ ceramics on molten steel, aluminum-killed steel was selected as the target melt, whose chemical composition is given in Table 2. The specific impregnation experiment was carried out as illustrated inFigure 3. Firstly, a 200 g steel block was placed in ZrO₂ crucible, and the porous ceramic was suspended above a block, then both were positioned together inside a graphite crucible. The assembly was placed in a vacuum induction furnace and heated to 1600 °C at a rate of 15 °C/min under an argon atmosphere. Subsequently, the porous ceramic was immersed in the melt and held for 20 mins when the steel was completely molten. Finally, the porous ceramic was removed and the molten steel was cooled in a furnace.

2.3. Characterization

The rheological property of the C₂M₂A₁₄ ceramic slurry at room temperature (25 ℃) was measured using a stress-controlled rheometer (Haake Mars40, Germany) under continuous shear mode. The phase composition of the specimens was characterized by an X-ray diffractometer (XRD, D8 DISCOVER A25, Germany) with the Cu-Kα radiation (test 2θ angle range: 10~80°, scanning speed: 5°/min). Scanning electron microscopy (SEM, NanoSEM 450, USA) equipped with an energy dispersive spectrometer (EDS, AMETEK EDAX, USA) was used to characterize the morphology and microstructure of the specimens. The bulk density and apparent porosity were determined via the Archimedes principle. The cold compressive strength (CCS) of the specimens was measured by the universal testing machine (ETM, MTS E45.105, America) according to the GB/T 1964-2023. Each performance was tested three times for the average value. The TSR of specimens was evaluated based on GB/T 16536-1996 via a water-cooling method [25,26]. The residual strength ratio (CCS after thermal shock testing/CCS before thermal shock testing) of the porous C₂M₂A₁₄ ceramics was calculated to evaluate its TSR. The thermal shock stability of the test specimens was averaged using three parallel specimens. The contents of [Al] and [O] in the steel were analyzed using an inductively coupled plasma spectrometer (ICP, Agilent 7800, America) and oxygen-nitrogen analyzer equipment (HORIBA EMGA-830, Japan). The inclusions quantity statistics in steel were examined with a fully automatic inclusion analyzer (ARL iSpark 8860, America). The removal efficiency of inclusions was evaluated based on Eq. 1 by comparing the T.O. before and after impregnation of molten steel.

(1)

where A and B represent the contents of T.O. in steel before and after impregnation using the porous C₂M₂A₁₄ ceramics, respectively, and F is the removal efficiency of inclusions in steel.

3. Results and Discussion

3.1. Physical Properties of Porous C₂M₂A₁₄ Ceramics

The rheological properties of the ceramic slurry critically influence the microstructure and performance of the final ceramic product. Figure 4a illustrates the rheological behavior of C₂M₂A₁₄ ceramics slurries containing different additions of ZrO₂ sol. From Figure 4a, the viscosity of all ceramic slurries decreases with increasing shear rate, exhibiting significant shear-thinning behavior. This characteristic facilitates rapid adhesion of the ceramic slurry to the PU foam scaffold surface during impregnation. Furthermore, the viscosity of the ceramic slurry increases with the increase of ZrO₂ sol addition, which is due to the enhanced interaction between the ZrO₂ sol network that increases the resistance to flow within the slurry. For all specimens, the slurry with 3 wt% ZrO₂ sol addition demonstrates the most favorable rheological properties. Figure 4b and Figure 4c show the bulk density, porosity and linear shrinkage ratio of as-prepared porous C₂M₂A₁₄ ceramics. Apparently, as the content of ZrO₂ sol increases from 0 to 3 wt%, the bulk density rises from 0.51 to 0.71 g/cm³, while the porosity decreases from 85.06 to 81.12% (Figure 4b). This phenomenon is mainly attributed to the fact that nano-sized ZrO₂ particles have an extremely high specific surface area and chemical activity, promoting the sintering of ceramic particles. However, as the ZrO₂ sol further increases to 4 wt%, the bulk density decreases to 0.46 g/cm³, which is due to the poor rheological properties and relatively high-volume expansion from the phase transformation of ZrO₂. The linear shrinkage rate shown in Figure 4c increases first and then decreases and the specimen ZS3 possesses the highest linear shrinkage of 13.55%.

3.2. Phase Composition and Microstructure

Figure 5 shows the phase composition of as-prepared porous C₂M₂A₁₄ ceramics with different ZrO₂ sol additions. For specimen ZS0, the C₂M₂A₁₄ is identified as the main phase. In contrast, for the porous C₂M₂A₁₄ ceramics containing various ZrO₂ sol additions (ZS2~ZS4), the C₂M₂A₁₄ and m-ZrO₂ phases are detected, indicating ZrO₂ sol has been successfully introduced into the porous C₂M₂A₁₄ ceramics. With the increase of ZrO₂ sol content, the intensity of m-ZrO₂ diffraction peak does not change significantly, which may be due to the relatively low addition of ZrO₂ sol.

Figure 6 presents the microstructure of as-prepared porous C₂M₂A₁₄ ceramics. For the specimen ZS0 (Figure 6a), the ceramic strut surface is composed of hexagonal plate-like grains and irregular particles packed together, with discernible interparticle pores. EDS result verified that these plate grains are mainly C₂M₂A₁₄. With the addition of ZrO₂ sol, as shown in Figure 6c~f, the strut becomes dense, which is attributed to the relatively high chemical activity of the nano-sized ZrO₂ particles, promoting the sintering of the strut. The EDS analysis results from spot 2 and spot 3 indicate the presence of Zr in addition to Ca, Mg, Al and O, verifying that ZrO₂ sol is successfully incorporated into the C₂M₂A₁₄. It can be seen that the bright white nano-sized ZrO₂ particles are uniformly dispersed on the plate-like C₂M₂A₁₄ grains. Meanwhile, the number of bright white ZrO₂ particles increases with increasing ZrO₂ sol addition. The elemental mapping of specimen ZS4 displayed in Figure 6i indicates that the elements Ca, Mg, Al and Zr are uniformly distributed in the scanned area, confirming the uniform dispersion of nano-sized ZrO₂ particles. However, when the ZrO₂ sol further increases to 4 wt%, some microcracks and pores appear on the strut surface (Figure 6g). This may be due to excessive volume expansion from the phase transformation, leading to more microcracks formation.

3.3. Thermal Shock Resistance

The impact of ZrO₂ sol addition on the CCS and TSR of porous C₂M₂A₁₄ ceramics is presented in Figure 7. From Figure 7a, the CCS firstly increased and then decreased with the increasing of ZrO₂ sol addition, which is consistent with the bulk density and porosity of specimens. The residual strength ratio (CCS after thermal shocks/CCS before thermal shocks) of porous C₂M₂A₁₄ ceramics firstly increases and then decreases when the ZrO₂ sol addition increases from 0wt% to 4wt%, and specimen ZS3 possesses the highest residual strength ratio of 66.4%. Figure 7b illustrates the TSR mechanism of as-prepared porous C₂M₂A₁₄ ceramics. Firstly, the hexagonal plate-like C₂M₂A₁₄ grains can facilitate crack deflection and dissipate crack energy during thermal shock. Furthermore, the phase transformation of uniformly distributed nano-sized ZrO₂ generates a few microcracks, which contribute to dispersing the propagation energy of the main crack. Meanwhile, owing to its high specific surface area, ZrO₂ sol promotes the sintering of the strut, improving the ability to resist thermal shock. These combined effects would lead to the excellent TSR of as-prepared porous C₂M₂A₁₄ ceramics. However, when the ZrO₂ sol addition reaches to 4 wt%, the excessive volume expansion during the phase transformation of ZrO₂ leads to relatively more cracks, which inevitably deteriorate its TSR.

Table 3 compares the CCS and TSR of specimen ZS3 with other porous ceramics reported in the literature [11,25,27,28,29,30]. It can be seen that when the porosity is comparable, specimen ZS3 exhibits a high compressive strength of 2.15 MPa and relatively high residual strength retention of 66.4%. This is attributed to the sintering promotion effect of ZrO₂ sol, the cross-stacked hexagonal platelet structures characteristics and the phase-transformation toughening effect of nano-sized ZrO₂ particles. Unlike irregular ceramic particles such as Al₂O₃, MgO, ZrO₂, SiC and MgAl₂O₄, the interlock hexagonal plate-like skeletal structure facilitates crack deflection and dissipates crack energy. Furthermore, the microcracks generated through the phase transformation between m-ZrO₂ and t-ZrO₂ are also beneficial for absorbing stress and dissipating crack energy [31,32,33,34].

3.4. Filter Performance

Considering both mechanical strength and TSR, specimen ZS3 is selected to evaluate its filter performance. Figure 8 displays the size and quantity of inclusions per 20 mm² scanned area in the steel before and after immersion tests using the specimen ZS3. It can be seen that the number of inclusions smaller than 1 μm, 1~3 μm, 3~5 μm and those larger than 5 μm in the impregnated steel is significantly reduced compared with those in unimpregnated steel. The removal efficiencies for those inclusions are calculated to be 91.25%, 71.45%, 81.44% and 91.07%, respectively, indicating an effective removal capability for various sizes of inclusions. Meanwhile, the content of Al and total oxygen also decreases markedly from 0.251 and 0.005 to 0.101 and 0.00158 after immersion with specimen ZS3 (Table 4), which proves that Al₂O₃ inclusions are effectively removed.

In order to clarify removal mechanism of the specimen ZS3, the cross-sectional microstructure of ZS3 before and after immersion is analyzed and the results are shown in Figure 9 and Table 5. Before immersion, the ceramic strut of specimen ZS3 consists of interlocked plate-like grains (Figure 9a~b), which are composed mainly of Ca, Mg, Al and O, with a small amount of Zr, confirming that they are ZrO₂-toughened C₂M₂A₁₄ (spot 1~2). After immersion, a reaction layer with an approximate thickness of 100 μm is formed (Figure 9c~d). For the reaction layer, the white particles primarily contain Fe (spot 3) and the elongated plate-like regions mainly contain Al, Ca and O, which correspond to the residual condensed Fe and mixtures of CaAl₁₂O₁₉ and CaAl₄O₇ (spot 4~5) inclusions. The irregular particles near the steel side consist mainly of Ca, Al, Mg and O, and are inferred to be mixtures of MgAl₂O₄, Al₂O₃, CaAl₄O₇ and CaAl₂O₄ (spot 6~8). This is consistent with the elemental mapping distribution in Figure 9e. Based on the above results, we infer that the removal mechanism of specimen ZS3 can be divided into two aspects: physical interception and chemical adsorption. For physical interception, the rough surface of the ceramic strut, created by the stacking of C₂M₂A₁₄, increases the contact area with molten steel, thereby enhancing the inclusions interception probability. For chemical-reaction adsorption, [Al] in molten steel first combines with [O] to form Al₂O₃ inclusions, subsequently reacts with C₂M₂A₁₄ composition to produce CaO-Al₂O₃ compounds and MgAl₂O₄, leading to the effective removal of Al₂O₃ inclusions.

4. Conclusions

This study systematically investigates the influence of ZrO₂ sol on the rheological properties, physical properties and thermal shock resistance as well as the inclusions removal mechanism of porous C₂M₂A₁₄ ceramics. The main conclusions can be drawn as follows:

(1)

The incorporation of highly active ZrO₂ sol promotes sintering, thereby enhancing compressive strength. Moreover, an appropriate amount of ZrO₂ sol improves the TSR by generating microcracks via phase transformation. These cracks can facilitate crack deflection and crack dissipation, which enhances thermal shock stability.

(2)

The optimized porous ZS3 ceramics exhibit a high compressive strength of 2.15 MPa and an excellent residual strength ratio of 66.4%. Owing to the synergistic effect of physical interception and chemical reaction, the as-prepared porous C₂M₂A₁₄-based ceramics achieve a high removal efficiency of 68.4% in total oxygen content. Given these superior properties, as-prepared porous C₂M₂A₁₄ ceramic is a promising candidate for molten-metal filtration applications.