The Damage Mechanism Of Magnesium-Chromium Refractories Used in Copper Smelting Side-Blowing Furnaces

 Side-blown furnaces are widely used in the smelting and processing of non-ferrous metals such as gold, silver, copper, lead, and tin. The refractory materials for the lining of side-blown furnaces need to have the characteristics of high temperature resistance, good thermal stability, high refractoriness under load, oxidation resistance, excellent resistance to high-temperature melt and melt erosion, and no participation in smelting slag formation. Based on the above performance requirements of side-blown furnaces for lining materials, magnesia refractories are commonly used as the refractory linings for side-blown furnaces. In order to improve the corrosion resistance of the lining materials, magnesium-chromium refractories with high Cr₂O₃ content are mostly used for the lining materials in parts in contact with copper matte melt and slag, such as the hearth and slag chamber. In recent years, domestic and foreign studies on the process of copper smelting furnaces and the damage mechanism of refractories have mostly focused on furnace types such as Ausmelt furnaces, Isa furnaces, bottom-blown furnaces, and converters, with fewer studies on the damage mechanism of magnesium-chromium refractories used in side-blown furnaces. In order to ensure the safe and stable operation of side-blown furnaces under new processes and improve the service life of lining materials, this paper analyzes the erosion and damage mechanism of used magnesia-chromite bricks in harsh working conditions of current side-blown furnaces, with a view to providing references for the quality improvement and innovation of magnesium-chromium refractories for side-blown furnaces.

1.1 Sample Preparation

Erosion behavior and mechanism analysis were conducted on used magnesia-chromite bricks that had been applied for 20 months in the slag line area of the side wall of a copper smelting plant's side-blown furnace. The used brick 6# from the slag line area and used brick 8# from the upper part of the slag line were selected as samples. Small blocks with dimensions of 40mm×20mm×5mm were cut by a small diamond cutting machine, cured with epoxy resin at 50℃ for 3 hours, and then ground and polished by a full-automatic pressure grinding and polishing machine. Samples including the slag coating layer, metamorphic layer (the densely structured part below the slag layer with significant macroscopic morphology difference from the original brick), and quasi-original brick layer (the part with small macroscopic morphology difference from the original brick) were prepared.

1.2 Detection and Characterization

The contents of SiO₂, Al₂O₃, Fe₂O₃, and K₂O in the smelted slag were detected according to GB/T6730.10-2014, GB/T6730.11-2007, GB/T6730.65-2009, and GB/T6730.49-2017, respectively. The contents of minor components were detected according to GB/T34333-2017. A Zeiss EVO-18 scanning electron microscope equipped with an X-ray energy spectrometer was used to analyze the microstructure and micro-area composition of the used magnesia-chromite brick samples.

2.1 Magnesium-Chromite Bricks Before Service

The apparent porosity and bulk density of electrofused rebonded magnesium-chromite bricks are 13.4% and 3.35 g·cm⁻³, respectively. The chemical composition (w) is as follows: MgO 55.42%, Cr₂O₃ 22.88%, Fe₂O₃ 12.29%, Al₂O₃ 6.34%, SiO₂ 0.74%, and CaO 0.76%. The morphological characteristics of the original bricks are shown in Figure 1. Figure 1(a) illustrates the bonding characteristics between particles and matrix of the magnesium-chromite bricks, where the aggregate and matrix are closely bonded. Figure 1(b) shows the phase distribution in the matrix, indicating that the matrix mainly includes chromite, periclase phase, and spinel phase precipitated within and between periclase crystals. Additionally, a small amount of forsterite phase (M₂S) is precipitated between periclase crystals.

Figure 1 Microstructure Photos of Original Magnesium-Chromite Bricks

2.2 Test Slag

The chemical composition (w) of the slag discharged from the side-blown furnace smelting is as follows: SiO₂ 27.05%, Al₂O₃ 4.90%, Fe₂O₃ 60.20%, CaO 2.20%, ZnO 1.99%, MgO 1.54%, CuO 1.10%, K₂O 1.02% (in this work, all chemical composition analysis results of Fe elements are expressed as Fe₂O₃, but iron elements in the ferrosilicon slag of the smelting environment mostly exist in the form of FeO). The mass ratio of Fe to SiO₂ is 1.53, and the mass ratio of CaO to SiO₂ is 0.08. The test shows that the refractoriness of the slag is 1380℃, the softening temperature of the slag is 1450℃, and the flow temperature is 1470℃.

2.3 Macroscopic Morphology of Used Magnesium-Chromite Bricks

The schematic diagram of the structural change of the side wall at the used slag line is shown in Figure 2. The 4# and 5# magnesium-chromite residual bricks are located at the lower part of the slag line, and there is no obvious slag coating on the surface of the residual bricks. The 7# and 8# magnesium-chromite residual bricks are located at the upper part of the slag line, and there is a thick slag coating layer on the surface of the residual bricks. The 6# magnesium-chromite brick is located at the interface between the slag and copper liquid in the slag line area. It can be seen that there is an obvious slag coating layer on the upper part of the slag line, while there is no obvious slag coating on the lower part of the slag line. The residual thickness of the used magnesium-chromite bricks and the thickness of the slag coating on the surface of the residual bricks are shown in Table 1. The residual thickness of the 7# and 8# residual bricks at the upper part of the slag line is greater than that of the 4#-6# residual bricks, and the thickness of the slag coating layer on the residual bricks at the upper part of the slag line also shows an upward trend with the increase of the distance from the slag line area. The residual thickness of the 6# residual brick at the slag line area is the smallest. The upper part of the slag line is in contact with the molten slag, which can form a slag coating layer on the surface of the magnesium-chromite bricks. Its erosion rate is lower than that of the magnesium-chromite bricks in contact with the copper liquid at the lower part of the slag line. However, due to the alternating scouring, penetration, and erosion of the copper liquid and molten slag in the slag line area, the 6# brick has the largest erosion rate.

Figure 2 Schematic Diagram of the Change of the Furnace Wall at the Slag Line of the Side - Blown Furnace After 20 - month Use

Table 1 Residual Thickness of Used Magnesium - Chromite Bricks and Thickness of Surface Slag Coating Layer

Figure 3 shows the cross-sectional photos of the used 6# and 8# magnesium-chromite residual bricks. By comparing the 6# and 8# residual bricks, the thickness of the dense metamorphic layer on their hot surfaces is both about 1 mm. It is speculated that under the working conditions here, due to frequent temperature fluctuations, the thickness loss of the hot surface of the magnesium-chromite bricks caused by scouring, thermal stress spalling, etc. is a continuous behavior, and a thick, obvious and continuous dense metamorphic layer will not be formed under the hot surface of the magnesium-chromite bricks. However, in most studies on the erosion and damage analysis of magnesium-chromite bricks for copper smelting furnaces, it has been found that a relatively thick metamorphic zone is formed and there are obvious stress cracks in the metamorphic zone [10-12]. That is, the damage mechanism of the magnesium-chromite bricks at the slag line of this side-blown furnace is significantly different from the damage mechanism of magnesium-chromite bricks under the furnace type and working conditions mentioned in the literature.

Figure 3 Cross-sectional Photos of Used Magnesium-Chromite Residual Bricks

2.4 Damage Analysis of Used Magnesium - Chromite Bricks

2.4.1 6# Magnesium - Chromite Residual Brick at the Lower Part of the Slag Line

Through comparative analysis of the microstructure and EDS composition of the 6# residual bricks located at the upper and lower parts of the slag line, it is found that although there are significant differences in the slag - coating conditions between the upper and lower parts of the slag line of the 6# magnesium - chromite residual brick, the microstructure, phase composition, and EDS element composition of the metamorphic layer and the quasi - original brick layer in the upper and lower parts of the slag line are similar. Therefore, only the microstructure, chemical composition, and phase composition of the 6# residual brick at the lower part of the slag line are used to analyze its erosion mechanism, and no repeated description is made for the upper part of the slag line. Figure 4 shows the microstructure photos at the hot surface of the 6# residual brick.

Figure 4 Microstructure photos of the hot surface of 6# residual brick

As shown in Figure 4, from top to bottom, the hot surface consists of a slag layer, a reaction layer (where the molten slag reacts with the original brick, causing significant changes in the composition and morphology of the matrix and particles), and a penetration layer (where the molten slag penetrates along pores and grain boundaries without obvious erosion reaction, and the composition and morphology of the matrix and particles remain unchanged). In the slag layer, the main phases are spinel and olivine-like phases [MFS: (Mg,Fe)₂SiO₄]. In the reaction layer, the magnesia and chromite in the original brick matrix undergo obvious dissolution and separation, generating a large amount of spinel and a small amount of olivine phases, accompanied by the penetration of Cu₂S and FeS, with a reaction layer thickness of about 1 mm. In the penetration layer, the periclase of magnesia and chromite particles are intact, and the molten slag erosion reaction is significantly weakened.

Figure 5 shows the microstructure photo of the reaction layer in Figure 4, and EDS analysis was conducted on its typical mineral phases, with the results shown in Table 2. Analysis of its chemical composition shows that the main phases are Cu₂S, FeS, aluminum-iron-chromium spinel, magnesium-iron olivine, and magnesium-aluminum-iron-chromium spinel. That is, at the hot surface of the 6# magnesium-chromite residual brick, the periclase phase in the original magnesium-chromite brick reacts with SiO₂ and FeO in the ferrosilicon slag to form low-melting-point magnesium-iron olivine, and the magnesia-chromite sand and chromite dissolve, absorbing Al₂O₃ and FeO in the ferrosilicon slag to generate iron-aluminum-chromium spinel and magnesium-iron-aluminum-chromium spinel.

Figure 5 Microstructure photos of the reaction zone of 6# residual brick

Figure 6 shows the microstructure photos of the 6# residual brick at different distances from the hot surface. It can be seen that a large amount of copper matte (bright white areas) are distributed from the hot surface to 35 mm away from the hot surface. The matrix pores and grain boundaries at 1-35 mm away from the hot surface are penetrated and filled by a large amount of copper matte, which wraps around the magnesia sand and chromite particles, resulting in the destruction of the bonding between magnesia-chromite sand, chromite, and periclase, and damaging the structural stability of the magnesia-chromite bricks.

Figure 6 Microstructure photos of different distances from the hot surface of 6# residual brick

EDS analysis shows that Cu element exists in the matrix of magnesia-chromite bricks only in the form of Cu₂S. Table 3 shows the comparison of chemical compositions of copper matte in the slag layer of 6# residual brick and at 1 mm and 20 mm away from the hot surface. It can be seen that the compositions of Cu, Fe and S elements in the copper matte at 1 mm and 20 mm remain basically unchanged, indicating that during the smelting process, the copper matte only undergoes penetration behavior in the magnesia-chromite bricks without obvious erosion reaction with the matrix and particles of magnesia-chromite bricks. Compared with the slag layer, the Cu content at 1 mm and 20 mm increases, while the Fe content decreases, which indicates that the penetration ability of Cu₂S in the copper matte in the magnesia-chromite bricks is stronger than that of FeS.

Table 3 Chemical compositions of the slag layer of 6# magnesium-chromite residual brick and the copper matte at 1 mm and 20 mm from the hot surface

Table 4 presents the EDS analysis results of the slag layer of the 6# residual brick and at different distances downward from the hot surface. It can be seen that: 1) The contents of SiO₂ and CaO at 1 - 35 mm from the hot surface are comparable. That is, during the copper smelting process, at the position of the 6# residual brick, the penetration of SiO₂ in the molten slag and its reaction with the magnesium-chromite brick only occur at the hot surface and do not further penetrate into the interior of the magnesium-chromite brick; 2) Compared with the chemical composition of the original brick, the contents of CuO and SO₃ at 1 - 35 mm increase significantly, which also indicates that the copper matte penetrates deeply into the interior of the magnesium-chromite brick.

Table 4 EDS Analysis Results of the Slag Layer of 6# Residual Brick and at Different Distances from the Hot Surface

Figure 7 shows the roadmap of the penetration of copper matte in the 6# residual brick. It can be seen that the copper matte mostly penetrates along the periphery of chromite particles and the intercrystalline regions where secondary iron-rich spinel precipitates. This phenomenon is related to the small wetting angle between the iron-rich chromite and the iron-rich spinel phases formed by secondary crystallization and the copper matte, which is more conducive to the penetration of copper matte in the magnesium-chromite brick. Therefore, high-purity magnesia, high-purity electrofused magnesia-chromite sand, and low-iron chromite are used to inhibit the precipitation of iron-rich mineral phases during the firing of the magnesium-chromite brick, improve the bonding strength of the magnesium-chromite brick, and reduce the degree of copper matte penetration. Some literatures show that under the same process conditions, improving the grades of magnesia and magnesia-chromite sand can enhance the performance of magnesia-chromite refractory materials, thereby improving their erosion resistance and permeability.

Figure 7 Penetration Route of Copper Matte in the Penetration Layer of 6# Residual Brick

From the above microstructure and EDS analysis of the 6# magnesium-chromite residual brick, at the slag line position: 1) SiO₂ and CaO in the ferrosilicon slag react with MgO dissolved from periclase on the hot surface to form a low-melting silicate phase. The penetration and erosion of magnesium-chromite bricks by SiO₂ and CaO can only continue after the hot surface of the magnesium-chromite brick is scoured and spalled; 2) At the same time, the copper matte shows penetration along pores and grain boundaries, and the penetration ability of Cu₂S in the copper matte is stronger than that of FeS; 3) Although the copper matte does not participate in the erosion reaction of the magnesium-chromite brick, the copper matte wraps around magnesia-chromite sand, chromite and magnesia sand particles, damaging the bonding degree of the magnesium-chromite brick, which is not conducive to the magnesium-chromite brick resisting the penetration and erosion of other high-temperature melts, thus accelerating the erosion and damage of the magnesium-chromite brick.

2.4.2 Magnesium-chromite Residual Brick at the Upper Part of the Slag Line

Figure 8 shows the microstructure photo at the hot surface of the 8# residual brick at the upper part of the slag line. Combined with the EDS surface scanning element distribution in Figure 9, different from the 6# residual brick, the 8# residual brick generates relatively large-sized (about 200 μm) iron olivine (F₂S) at the hot surface; below the iron olivine is a dense iron-rich spinel layer (with a mass fraction of Fe₂O₃ being 89.8%), and below the spinel layer is a dense magnesium-iron olivine (MFS) layer. This indicates that the hot surface of the magnesium-chromite brick at the upper part of the slag line is mostly a relatively stable molten slag, providing the thermal and dynamic conditions conducive to the reaction between the molten slag and the magnesium-chromite brick to form olivine and a dense layer of secondary spinel. The spinel layer and the magnesium-iron olivine layer have a dense structure, which can play a certain role in preventing the further penetration and erosion of molten slag and copper matte, delaying the degree of structural and compositional changes of the magnesium-chromite brick, and reducing the structural spalling rate of the magnesium-chromite brick.

Figure 8 Microstructure photos of the hot surface of 8# residual brick

Figure 9 EDS surface scanning element distribution map of the hot surface of 8# residual brick (Figure 8)

Figure 10 shows the microstructure photos of the 8# residual brick at different distances from the hot surface. As can be seen from Figure 10, compared with the microstructure of the 6# residual brick: 1) Although the 8# residual brick is located at the upper part of the slag line, copper matte still penetrates into the 8# residual brick, but the penetration degree of copper matte in the 8# residual brick is significantly reduced. With the increase of the distance from the hot surface, the amount of pores not filled by copper matte gradually increases. Table 5 shows the EDS analysis results of the 8# residual brick at different distances from the hot surface. Compared with the copper matte content of the 6# residual brick at different distances from the hot surface, it is obvious that the copper matte content at the same hot surface distance in the 8# residual brick is significantly reduced. 2) The reaction degree between the molten slag and the magnesium-chromite brick in the reaction layer is weaker than that of the 6# residual brick. In the reaction layer of the 8# residual brick, part of the chromite only undergoes decomposition and secondary spinelization reactions at the particle edges, and the particle boundaries are still relatively clear.

(1) The erosion rate of magnesia-chromite bricks in the slag line area is higher than that in the upper part of the slag line. Under the repeated scouring, penetration, and erosion of copper matte melt and SiO₂-FeO slag, the magnesia-chromite bricks in the slag line area cannot form a stable slag coating layer, resulting in the fastest erosion rate.

(2) In the cross-sections of magnesia-chromite residual bricks in both the slag line and upper slag line areas, only a metamorphic layer of approximately 1 mm is observed below the hot surface. It is inferred that the metamorphic layer formed by slag penetration and erosion during the service of magnesia-chromite bricks cannot be stably maintained at the hot surface. The bricks are subjected to penetration and erosion reactions by copper matte melt and SiO₂-FeO slag, followed by hot surface spalling. Therefore, to improve the service life of magnesia-chromite bricks, it is feasible to enhance the density and bonding degree of magnesia-chromite materials to strengthen their anti-penetration and anti-erosion properties, reduce the thickness of the metamorphic layer, or stabilize the smelting process to minimize temperature fluctuations, thereby reducing thermal spalling.

(3) Both the lower and upper parts of the slag line magnesia-chromite residual bricks experience copper matte penetration and SiO₂-FeO slag erosion, and the reaction between SiO₂-FeO slag and magnesia-chromite bricks generates low-melting olivine phases and secondary spinel phases. However, the internal penetration of copper matte in the slag line area is more severe. After penetrating, copper matte wraps around matrix particles such as chromite, destroying the bonding strength of magnesia-chromite bricks and significantly reducing their resistance to thermal shock and spalling from copper matte melt and SiO₂-FeO slag. In the upper part of the slag line, the penetration of copper matte is weakened, with SiO₂-FeO slag erosion being the main hot surface reaction. A dense secondary spinel layer and magnesium-iron olivine layer form at the hot surface, which can reduce the penetration of slag and copper matte melt into magnesia-chromite bricks, mitigate the structural damage caused by SiO₂-FeO slag and copper matte, and slow down the erosion rate of magnesia-chromite bricks.

(4) Copper matte mainly penetrates along the grain boundaries and pores around chromite, as well as the intercrystalline regions of periclase where iron-rich spinel precipitates, which is detrimental to the resistance of magnesia-chromite bricks to the penetration and erosion of SiO₂-FeO slag. Measures such as improving raw material purity, reducing the FeO content in chromite and magnesia-chromite sand, and inhibiting the precipitation of Fe-bearing phases can be adopted to reduce the penetration of copper matte into the interior of magnesia-chromite bricks, thereby decreasing the erosion rate of magnesia-chromite bricks.